气象站观测数据插补以及垂直观测风廓线缺测数据插补

为揭示复杂地形条件下云南省年降水量的空间分布特征，并筛选适用于区域尺度研究的最优空间插值方法，本文基于云南省27个气象站点2022年年降水观测数据，采用反距离加权法(Inverse Distance Weighting, IDW)、普通克里金法(Ordinary Kriging, OK)、径向基函数法(Radial Basis Function, RBF)和趋势面插值法(Trend Surface Interpolation, Trend)对年降水量进行空间插值模拟。通过对不同插值结果空间分布特征的对比分析，并结合交叉验证方法对插值精度进行定量评价，系统评估各方法的适用性。结果表明，不同插值方法在宏观尺度上均呈现出“滇西南高、滇东北低”的降水空间分异特征，但在空间连续性和平滑性方面存在显著差异。其中，RBF方法在保持区域地带性特征的同时，能够有效避免局地异常值对插值结果的过度影响，插值结果空间分布最为合理。综合精度指标与空间表达效果，RBF插值方法在云南省复杂地形条件下表现出最佳的综合适用性。

普通克里金法为一种广泛使用的地理统计的插值方法，但一般都只依据经验使用一个变异函数来计算插值结果。同时，现有研究着重研究年、月尺度下普通克里金法不同变异函数的比较，对日降水尺度研究较少。为了得到普通克里金法在日降水插值适应性较好的变异函数，比较和分析普通克里金法四种常用不同变异函数(指数、球型、高斯、线性)。选取了湘江洣水所在子流域及子流域周边一共43个雨量站1990~2005年逐日降水资料插值，从相关系数、检验指标、不同降水级别准确率这三方面比较插值结果，交叉验证的结果表明：1) 指数和球型函数对普通克里金法的日降水插值适应性较好；2) 普通克里金法的插值结果在降水量大的日子普遍偏小；3) 随着降水量级的增大，日降水插值的误差明显增大。

由于观测手段和观测背景的限制，再加上环境复杂，很多时候只有部分气象观测资料可用，为了在这种背景下进行气象预报，充分完备的气象资料是重要基础，因此基于零散的部分观测数据、先验数据的统计特征和矩阵优化填充技术的气象资料恢复研究具有重要的工程价值和数学意义，其研究在国内外尚属空白。本文旨在通过部分观测资料，充分利用矩阵的低秩性和气象观测数据的内在结构性先验统计信息，应用矩阵填充的奇异值阈值化SVT算法，优化分析得到欠缺数据，从而获得填充的补全数据。实验结果表明，基于结构性先验信息和矩阵优化填充方法得到的数据准确率明显取决于矩阵格式选择和气象数据本身特性，而且本文通过理论和实验分析出最佳的矩阵优化填充模型，表明当可利用的资料占比高于临界采样率时，数据填充误差可控制在10%以内，可以有效地解决预报和分析时的观测资料数据缺失不全的问题。

本文深入探讨了ArcGIS在气象业务中的多元应用。通过详细阐述ArcGIS在气象数据管理、数据可视化、空间分析以及灾害管理等方面的具体运用，展现了该软件在提升气象数据处理效率、增强数据可视化效果、深入挖掘气象数据价值以及优化气象灾害管理方面的显著优势。文章首先介绍了ArcGIS在气象数据管理中的应用，包括数据采集、整合、存储、质量控制、查询更新及数据共享等方面。随后，文章讨论了ArcGIS在气象数据可视化中的重要作用，如数据导入、空间插值、可视化效果展示及结果输出等。在气象空间分析方面，ArcGIS提供了丰富的工具和功能，有助于研究人员深入挖掘气象数据的内在规律和趋势。最后，文章强调了ArcGIS在气象灾害管理中的关键作用，包括灾害风险评估、监测预警、应急响应及灾后重建等方面。通过ArcGIS的综合应用，气象部门能够更高效地处理和分析气象数据，为相关部门和公众提供更优质的气象服务和决策支持。

气象设备的稳定运行对航空安全与效率至关重要，而传统维护模式面临设备老化、人工经验依赖及多源数据整合不足等挑战。本研究提出基于Hadoop与Spark技术的智能化维护解决方案：通过模块化传感器组网、边缘计算节点及Kafka消息队列实现数据高效采集与实时同步；运用3σ准则、滑动窗口算法、插值法及NTP协议完成数据清洗与标准化；从原始数据中提取设备健康指标及时序特征，结合环境变量通过Spark MLlib训练LSTM故障预测模型；最终通过动态预警阈值、工单智能生成及三维态势可视化实现维护决策优化。研究成果显著提升了维护效率与准确性，未来可结合边缘计算与跨域数据融合技术进一步优化。

首先以考虑地形遮蔽的分布式可照时间理论模型结合气象台站观测的日照时间计算了日照百分率，然后对我国第二代静止气象卫星(FY-2G)的总云量遥感影像产品重采样后，根据日照百分率和云量的负相关性，分别建立了基于遥感总云量和观测站云量(总云量、低云量)的日尺度日照百分率估算模型，并以气象站点分布稀疏的新疆区域为例，对两种估算模型使用IDW和Kriging插值方法的模拟效果进行了检验，得出以下结论：1) 实际地形下新疆区域可照时间的空间分布受地形影响较大，全区四季平均可照时间分别为：春季1165 h，夏季1286 h，秋季964 h，冬季823 h。2) 基于遥感云量的条带状重采样方案考虑了日照轨迹和云的区域移动变化，重采样后的云量值与日照百分率的相关性有所提升，相关系数为0.756。3) 建立的单站分季节遥感集成日照百分率模型相关系数冬夏差异明显，夏季最高，春秋次之，且分布形态较为一致，冬季最低，低值主要集中在北疆沿天山一带。4) 从模拟的效果来看，遥感集成日照百分率模型(平均绝对误差为14.8%)要明显优于基于观测站云量模型模拟的结果，由于遥感集成日照百分率充分发挥了卫星在空间上连续观测的优势，空间分布更为连续，通过检验后的该方法可以在站点稀少的西部地区进行业务应用。

浙江地区汛期雷暴大风天气多发频发，常导致经济损失和人员伤亡。为提高短临预报水平，文章利用地面自动气象站、多普勒天气雷达、风廓线雷达资料，选取2022年浙北地区雷暴大风个例，探究了雷达逆风区与边界层大风的关系并计算了地面大风的预警提前量。结果表明：多普勒天气雷达能提前探测到中低空和边界层逆风区的发生发展，较地面大风出现有平均16.9 min和11.4 min的提前量。通过2020~2022年的典型强对流大风个例分析，多普勒雷达逆风区–风廓线雷达边界层大风–自动站地面大风三者之间能形成短临预报提前量时间链，对地面雷暴大风的短临预警有指示意义。此外，2022年7月12日发生的下击暴流事件中，多普勒雷达最低仰角探测水平风场变化不明显，但风廓线雷达探测到高空风速突然增大，早于地面大风出现时间约33 min，表明风廓线雷达有利于提高下击暴流的提前预报预警能力。

热岛效应是指由于城市化所引起的城市地表及大气温度高于周边郊区的现象，是城市气候最显著的特征之一。台站气温资料一直被作为研究城市热岛时空演变规律的重要手段，随着观测资料的不断发展，更加精细化的自动站气温资料也被用于研究城市热岛的更多细节特征。本文即基于加密台站逐小时气温资料，利用克里金插值方法计算格点气温数据，结合土地利用/土地覆盖数据划分城市郊区范围，分析研究厦门精细化城市热岛的24小时时空演变特征。结果表明，热岛强度夜间大于白天，在00时热岛效应影响范围最大，主要分布在厦门市岛内北部和岛外的环岛区域的低温区。

为了提高GRAPES模式下的温度产品的准确性，本文采用2021年全年GRAPES模式下的温度数据作为预报数据，欧洲中心的ERA产品作为实况数据，数据点分布为格点分布。研究区域为30.61˚N~30.73˚N，103.78˚E~103.90˚E，将该范围内的25个网格点上的数据，按照每1个小时一次的时间间隔，通过Kriging插值处理预报数据的初值场插值到靠近温江站的格点30.7˚N，103.8˚E，得到较为稳定的预报数据后，运用线性拟合从总体距平值最小的角度分析最为符合预报–实况温度的拟合线；运用三次样条插值拟合从点与点之间的斜率关系分析最为符合预报和实况温度的拟合线；还运用神经网络分析训练，以尽量减小预报温度和实况温度间的异常大的差为目的，来拟合最符合预报和实况温度的曲线。对于订正结果，分析对比订正后拟合结果与实际观测数据的残差，来判定某一订正方法的可信度高低。结果表明：1) 经过Kriging插值处理过后的初值场，残差数值大小从整体上已经呈现出小值多，大值少，和实况数据较为贴近。2) 三种订正方法的拟合结果，均比原来Kriging插值的初值场更加贴近实况数据，残差数值小的区域均数值点更加密集。3) 三种订正方法对比，三次样条插值拟合的残差数值大的区域有较多的拟合数据点，尤其是残差超过5℃的区域，容易引起部分时间点，拟合和实况有较大出入；线性拟合与神经网络分析训练的拟合结果的残差数值小的区域数据分布均很稀疏，但是二者对比，神经网络分析训练的拟合结果，拟合数据与实况数据的残差分布大半都集中在−2℃~2℃的小值区域，结果与实况大致相同。

空间数据插值算法是目前普遍应用的一种科学算法，其广泛应用于气象、农业及地质勘探等领域。通过空间数据插值，对缺省或非有效空间数据进行估计和推测，在投资较少的情况下得到大量精度能够满足研究及生产需要的数据，可以在很大程度上满足人类的需要，具有很强的现实意义。本文详细介绍了两种常用的空间数据插值算法——反距离权重插值算法(IDW)和普通克里金插值算法(OK)的原理，并且基于对反距离权重插值算法的研究，针对其可优化参数——距离幂指数k，提出了一种改进的反距离权重插值法。通过数值模拟对济南市的空气质量状况进行空间插值分析，可以得出结论：(1) 普通克里金插值算法优于反距离权重插值算法；(2) 在普通克里金插值算法中，指数模型的插值效果较好；(3) 本文提出的改进的反距离权重插值算法较原算法的插值效果更佳。

利用1970~2019年疏勒河与黑河流域共12个地面观测站的气温、降水等逐日观测资料，采用一元线性回归、克里金插值、Mann-Kendall法等方法对疏勒河和黑河流域气候变化特征进行分析，结果表明：1) 两流域气温和降水的时间变化均呈上升趋势。2) 两流域气温和降水空间分布分别呈西高东低、东多西少的特征。3) 疏勒河流域气温和降水量均与海拔呈负相关、与经度和纬度呈正相关；黑河流域气温与海拔呈负相关、与经度和纬度呈正相关，降水量与海拔和经度呈正相关、与纬度呈负相关。4) 近50年来，两流域气温发生突变的时间均为1994年；疏勒河流域降水量突变开始时间为1970和2019年，黑河流域为2012年。5) 两流域的极端高温指数呈上升趋势，持续干燥指数呈显著下降趋势；疏勒河流域容易发生极端高温事件、出现旱情，黑河流域更容易发生极端低温和极端降水事件。

改善城市内部的空气质量(如雾霾、臭氧、二氧化硫等)可通过建设城市通风廊道来实现。本文通过地质统计学的普通克里金插值方法，基于北京市PM2.5浓度观测值，代表性研究东西、南北、西南–东北、东南–西北四个方向上潜在的北京市通风廊道建设线路。综合考虑地形特点，结合不同的地形和风向，使通风潜力最大化。最终给出了准确的四条通风廊道的线路走向，发现此四条线路基本均经过顺义区、海淀区、通州区。为日后城市通风廊道规划与建设提供了一些参考。

本文利用贵阳机场CFL-03型风廓线雷达结合多源数据对2021年7月17日贵阳机场一次强对流天气进行了综合分析，研究表明：风廓线雷达资料能很好地反映贵阳机场上空水平风随高度的分布情况，从而为强对流天气的短临监测和预警提供十分有用的帮助信息，诸如判断低层大气冷暖平流，低空急流强度，冷空气入侵信号等；通过风廓线雷达的实时监测和产品应用，可以有效地捕捉到强对流天气发生前本地边界层内大气环境发生变化的微弱讯号，风廓线雷达的垂直速度、信噪比和谱宽都与短时强降水有着很好的关联，且信噪比和谱宽在强降水开始前30分钟左右出现了显著的变化特征，这些信号再结合多普勒气象雷达的实时探测，可以为高时空分辨率的预警服务提供有力的支撑；贵阳机场风廓线雷达安装于跑道南端入口处，其不仅可以实时测风，还可以计算出水平风的垂直切变指数，从而为进近和起飞的飞机提供垂直方向上的风切变预警。

本文利用EC再分析资料ERA5 (0.25˚ × 0.25˚)、风廓线雷达资料、自动气象观测站资料、探空资料，对2024年5月19日南宁吴圩国际机场发生的强对流天气过程进行分析。结果表明：(1) 高空短波槽东移、中低层低涡切变系统活动与地面冷空气南下共同作用，是此次强对流天气的主要触发条件；(2) 低层湿舌北伸与偏南暖湿气流持续输送，为对流发展提供了充足的水汽供应和辐合抬升条件；(3) 南宁机场探空“上干下湿”的垂直结构，有利于对流不稳定能量积累与释放；(4) 低层辐合与高层辐散形成的垂直抽吸效应，是此次对流天气维持、发展并增强的关键动力机制。

Univariate imputation methods are defined as imputation methods that only use the information of the target variable to estimate missing values. While univariate imputation methods are convenient and flexible since no other variable is required, multivariate imputation methods can potentially improve imputation accuracy given that the other variables are relevant to the target variable. Many multivariate imputation methods have been proposed, one of which is Vector Autoregression Imputation Method (VAR-IM). This study aims to compare imputation results of VAR-IM-based methods and univariate imputation methods on time series data, specifically on long lag seasonal data such as daily weather data. Three modified VAR-IM methods were studied using simulations with three steps: deletion, imputation, and evaluation. The deletion step was conducted using six different schemes with six missing proportions. The simulations were conducted on secondary daily weather data collected from meteorological station of Citeko from January 1, 1991, to June 22, 2013. Nine weather variables were examined, that is the minimum, maximum, and average temperatures, average humidity, rainfall rate, duration of solar radiation, maximum and average wind speed, as well as wind direction at maximum speed. The simulation results show that the three modified VAR-IM methods can improve the accuracies in around 75% of cases. The simulation results also show that imputation results of VAR-IM-based methods tend to be more stable in accuracy as the missing proportion increase compared to the imputation results of univariate imputation methods.

Meteorological records, including precipitation, commonly have missing values. Accurate imputation of missing precipitation values is challenging, however, because precipitation exhibits a high degree of spatial and temporal variability. Data-driven spatial interpolation of meteorological records is an increasingly popular approach in which missing values at a target station are imputed using synchronous data from reference stations. The success of spatial interpolation depends on whether precipitation records at the target station are strongly correlated with precipitation records at reference stations. However, the need for reference stations to have complete datasets implies that stations with incomplete records, even though strongly correlated with the target station, are excluded. To address this limitation, we develop a new sequential imputation algorithm for imputing missing values in spatio-temporal daily precipitation records. We demonstrate the benefits of sequential imputation by incorporating it within a spatial interpolation based on a Random Forest technique. Results show that for reliable imputation, having a few strongly correlated references is more effective than having a larger number of weakly correlated references. Further, we observe that as the proportion of stations with incomplete records increases, there is a higher percentage of stations that benefit from sequential imputation. Overall, we present a new approach for imputing missing precipitation data which may also apply to other meteorological variables.

Abstract Representativeness and quality of collected meteorological data impact accuracy and precision of climate, hydrological, and biogeochemical analyses and predictions. We developed a comprehensive Quality Assurance (QA) and Quality Control (QC) statistical framework, consisting of three major phases: Phase I—Preliminary data exploration, i.e., processing of raw datasets, with the challenging problems of time formatting and combining datasets of different lengths and different time intervals; Phase II—QA of the datasets, including detecting and flagging of duplicates, outliers, and extreme data; and Phase III—the development of time series of a desired frequency, imputation of missing values, visualization and a final statistical summary. The paper includes two use cases based on the time series data collected at the Billy Barr meteorological station (East River Watershed, Colorado), and the Barro Colorado Island (BCI, Panama) meteorological station. The developed statistical framework is suitable for both real-time and post-data-collection QA/QC analysis of meteorological datasets.

Ground meteorological observation data (GMOD) are the core of research on earth-related disciplines and an important reference for societal production and life. Unfortunately, due to operational issues or equipment failures, missing values may occur in GMOD. Hence, the imputation of missing data is a prevalent issue during the pre-processing of GMOD. Although a large number of machine-learning methods have been applied to the field of meteorological missing value imputation and have achieved good results, they are usually aimed at specific meteorological elements, and few studies discuss imputation when multiple elements are randomly missing in the dataset. This paper designed a machine-learning-based multidimensional meteorological data imputation framework (MMDIF), which can use the predictions of machine-learning methods to impute the GMOD with random missing values in multiple attributes, and tested the effectiveness of 20 machine-learning methods on imputing missing values within 124 meteorological stations across six different climatic regions based on the MMDIF. The results show that MMDIF-RF was the most effective missing value imputation method; it is better than other methods for imputing 11 types of hourly meteorological elements. Although this paper applied MMDIF to the imputation of missing values in meteorological data, the method can also provide guidance for dataset reconstruction in other industries.

In this work, we used the MICE (Multivariate Imputation by Chained Equations) technique to impute missing daily data from six meteorological variables (precipitation, temperature, relative humidity, atmospheric pressure, wind speed and insolation) from 96 stations located in the northeast region of Brazil (NEB) for the period from 1961 to 2014. We then applied tests with a quality control system (QCS) developed for the detection, correction and possible replacement of suspicious data. Both the applied gap filling technique and the QCS showed that it was possible to solve two of the biggest problems found in time series of daily data measured in meteorological stations: the generation of plausible values for each variable of interest, in order to remedy the absence of observations, and how to detect and allow proper correction of suspicious values arising from observations.

Abstract Modeling by multiple enchained imputation is an area of growing importance. However, its models and methods are frequently developed for specific applications. In this study the model for multiple imputation was used to estimate daily rainfall data. Daily precipitation records from several meteorological stations were used, obtained from system AGRITEMPO for two homogenous climatic zones. The precipitation values obtained for two dates (Jan. 20th 2005 and May 2nd 2005) using the multiple imputation model were compared with geo-statistics techniques ordinary Kriging and Co-kriging with the altitude as an auxiliary variable. The multiple imputation model was 16% better for the first zone and over 23% for the second one, compared to the rainfall estimation obtained by geo-statistical techniques. The model proved to be a versatile technique, presenting coherent results with the conditions of different zones and times.

No abstract

The availability of solar irradiance time series without missing data is an ideal scenario for researchers in the field. However, it is not achievable for a variety of reasons, such as measurement errors, sampling gaps, or other factors. Time series imputation methods can be a solution to the lack of data and, in this paper, we study the applicability of Bidirectional Encoder Representations from Transformers (BERT) as an irradiance time series imputation solution. In this regard, a BERT model was trained from scratch for the masked language modelling (MLM) task, and the quality of the imputation was evaluated according to the number of missing values and the position within the series. The experiments were conducted over a dataset of 165 stations, captured by meteorological stations distributed over the Spanish regions of Galicia, Castile, and León. In the evaluation process, an average coefficient of determination (R2 score) of 0.89% was obtained, the maximum result being 0.95%.

Missing rainfall data have reduced the quality of hydrological data analysis because they are the essential input for hydrological modeling. Much research has focused on rainfall data imputation. However, the compatibility of precipitation (rainfall) and non-precipitation (meteorology) as input data has received less attention. First, we propose a novel pre-processing mechanism for non-precipitation data by using principal component analysis (PCA). Before the imputation, PCA is used to extract the most relevant features from the meteorological data. The final output of the PCA is combined with the rainfall data from the nearest neighbor gauging stations and then used as the input to the neural network for missing data imputation. Second, a sine cosine algorithm is presented to optimize neural network for infilling the missing rainfall data. The proposed sine cosine function fitting neural network (SC-FITNET) was compared with the sine cosine feedforward neural network (SCFFNN), feedforward neural network (FFNN) and long short-term memory (LSTM) approaches. The results showed that the proposed SC-FITNET outperformed LSTM, SC-FFNN and FFNN imputation in terms of mean absolute error (MAE), root mean square error (RMSE) and correlation coefficient (R), with an average accuracy of 90.9%. This study revealed that as the percentage of missingness increased, the precision of the four imputation methods reduced. In addition, this study also revealed that PCA has potential in pre-processing meteorological data into an understandable format for the missing data imputation.

Abstract Rainfall forecasting model using data Global Circular Model (GCM) with Statistical Downscaling technique has a fairly high accuracy. However, missing local climate information poses a constraint in data analysis and forecasting. Missing value imputation is one solution that can be used. Kalman Filter Imputation and State Space Model Arima are imputation methods that operate recursively where there is an update of prediction values when data updates occur. This study aimed to find the best model to use for missing value imputation with small imputation errors. The results of the missing value imputation were used to obtain the best statistical downscaling model on a 3 × 3 to 12 × 12 grid. The research was conducted on the daily rainfall data of Kupang City with 17% missing values and 8% unusual data at the Eltari Meteorological observation station, Kupang city. The average daily rainfall data in East Nusa Tenggara Province were utilized as a reference for the characteristics of rainfall data at the Kupang City observation station. The best missing value imputation was obtained by using the Arima State-Space Model (2,1,1) with a Root Mean Square Error (RMSE) of 0.930 and the model was statistical downscaling best obtained on a grid 6 × 6 with a Mean Absolute Percentage Error (MAPE) of 1.3 % and the number of PCs 11.

No abstract

With the ultimate goal of developing models that involve the use of environmental variables, a GIS-based application is being developed that is circumscribed to the region of Galicia (Spain). Ten-minute data of six meteorological variables were collected from 150 stations of the MeteoGalicia network over a period of 18 years, but the time series data are not complete. In order to estimate missing rainfall data, four imputation methods were evaluated in this study: missForest, MICE, Amelia II, and inverse distance weighting (IDW). Crossvalidation results show that the precipitation is out of phase in the different stations due to their geographical locations, and the imputation can be improved with a displacement of the time series; on the other hand, the missForest method provided better results in the imputation of this meteorological variable than the MICE, Amelia, or IDW.

No abstract

The presence of missing rainfall data is inevitable due to error of recording, meteorological extremes and malfunction of instruments. Consequently, a competent imputation algorithm for missing data treatment algorithm is very much needed. There are several such efficient algorithms which have been introduced in earlier studies. However, the limitations of current algorithms are they are highly dependent on the information and homogeneity of adjoining rainfall stations. Therefore, this study is intended to introduce several single imputation algorithms for missing data treatment, which believed to be more competent in treating missing daily rainfall data without the need to depend on the information of adjoining rainfall stations. The proposed algorithms use descriptive measures of the data, including arithmetric means, geometric means, harmonic means, medians and midranges. These algorithms are tested on hourly rainfall data records from six selected rainfall stations located in the Kuantan River Basin. Based on the analysis, the proposed singular imputation algorithms, which treated missing data by geometric means, harmonic means and medians are more superior compared to the other imputation algorithms, irrespective of missing rates and rainfall stations.

No abstract

The adoption of climate-smart agriculture requires the comprehensive development of environmental monitoring tools, including online observation of climate and soil settings. They are often designed to measure soil properties automatically at different depths at hour or minute intervals. It is essential to have a complete dataset to use statistical models for the prediction of soil properties and to make short-term decisions regarding soil tillage operations and irrigation during a vegetation period. This is also important in applied hydrological studies. Nevertheless, the time series of soil hydrological measurements often have data gaps for different reasons. The study focused on solving a problem of gap-filling in hourly time series of soil temperature and moisture, measured at the 30 cm depth using a weighted gravitation lysimeter station while meteorological data were recorded simultaneously by a weather station. The equipment was installed in the Kulunda Steppe in the Altai Krai, Russia. Considering that climate conditions affect soil temperature and moisture content directly, we did a comparative analysis of the gap-filling performance using the three imputation methods—linear interpolation, multiple linear regression, and extended ARMA (p,q) models with exogenous climatic variables. The results showed that, according to the minimum of the mean absolute error, ARMA (p,q) models with optimally selected order parameters, and an adaptive window, had some advantages compared to other single-imputation methods. The ARMA (p,q) model produced a good quality of gap-filling in time series with the mean absolute error of 0.19 °C and 0.08 Vol. % for soil temperature and moisture content, respectively. The findings supplemented the methodology of hydrological data processing and the development of digital tools for the online monitoring of climate and soil properties in agriculture.

Computational methods based on machine learning have had extensive development and application in hydrology, especially for modelling systems that do not have enough data. Within this problem, there are data series that are missing, and that should not necessarily be discarded; this is achieved by means of the imputation of the same ones, obtaining complete sets. For this reason, this research proposes a comparison of computer-learning techniques to identify those best suited for hydrographic systems of the Pacific of Ecuador. For the elaboration of this investigation, the hydro-meteorological records of the monitoring stations located in the watersheds of the Esmeraldas, Cañar and Jubones Rivers were used for 22 years, between 1990 and 2012. The variables that were imputed were precipitation and flow. Automatic learning machines of the Python Scikit_Learn module were used; these modules integrate a wide range of automated learning algorithms, such as Linear Regression and Random Forest. Finally, results were obtained that led to a minimum useful mean square error for Random Forest as an automatic machine-learning imputation method that best fits the systems and data analyzed.

Abstract The missing of the meteorological data in Iraq is common due to malfunction of measuring devices, security status, and human effects. The study tested 17 missing precipitation data estimation methods in Baghdad city as a case study, where, all the surrounding stations around Baghdad experienced the missing of data for various reasons, and some of the missing data are for a full year record. The methods examined in this study are based on different approaches, some of the methods are based upon the distances to the targeted station, others are upon regression factors, and there are also methods that combine several factors. There are also other types of missing data filling methods which depend on imputation and artificial intelligence. The investigation of the most accurate method to find the missing data will assist researchers and decision makers to fill the gap in their analysis in one of the most vulnerable countries in terms of drought and climate changes impacts. Results showed that Expectation Maximization (EM) method utilization has the best results with the least errors, and Multiple Linear Regression (MLR) method was ranked the second best method. In general, all of the applied methods had resulted acceptable interpolations, and it was clear that the combined methods have low significance on the results in comparison with others. All of these findings are limited to the study area meteorological and spatial conditions.

No abstract

No abstract

Addressing data incompleteness issues is crucial for reliable climate studies, especially in regions like Africa that commonly experience data gaps. This study aims to evaluate the performance of five imputation methods (knn, ppca, mice, imputeTS, and missForest) on meteorological data from stations in Burkina Faso and Senegal. The imputed data is compared with ERA5 reanalysis data to validate its accuracy. Temperature, relative humidity, and precipitation observations from the GSOD dataset (1973-2020) were used, creating subsets with missing rates of 5, 10, 20, 30 and 40%. An evaluation was conducted using the Taylor diagram and Kling-Gupta Efficiency (KGE). The results show a good estimation of temperature and relative humidity time series, with missForest performing the best for handling missing values. Precipitation estimation was less accurate, but there was strong agreement between estimated and observed data. ImputeTS was recommended for precipitation. Spatial consistency between imputed data and ERA5 reanalysis products was found. This research improves the quality of meteorological data, provides essential information about climatic characteristics, and serves as a foundation for climate change and weather modeling studies.   Key words: Meteorological data, imputation methods, Senegal, Burkina Faso.

The analysis of rainfall data with no missingness is vital in various applications including climatological, hydrological and meteorological study. The issue of missing data is a serious concern since it could introduce bias and lead to misleading conclusions. In this study, five imputation methods including simple arithmetic average, normal ratio method, inverse distance weighting method, correlation coefficient weighting method and geographical coordinate were used to estimate the missing data. However, these imputation methods ignored the seasonality in rainfall dataset which could give more reliable estimation. Thus this study is aimed to estimate the missingness in daily rainfall data by using generalized linear model with gamma and Fourier series as the link function and smoothing technique, respectively. Forty years daily rainfall data for the period from 1975 until 2014 which consists of seven stations at Kelantan region were selected for the analysis. The findings indicated that the imputation methods could provide more accurate estimation values based on the least mean absolute error, root mean squared error and coefficient of variation root mean squared error when seasonality in the dataset are considered.

Accurate and reliable wind speed data are essential across diverse wind engineering applications, such as maximizing the efficiency and effectiveness of wind energy utilization. However, the continuity of wind speed monitoring is often disrupted by missing data due to sensor malfunctions, adverse environmental conditions, or limited measurement coverage. To address this challenge, this study introduces the CNN-based Wasserstein Generative Adversarial Imputation Network (C-WGAIN), a novel framework designed to impute missing wind speed data. The proposed framework employs two-dimensional convolutional layers to capture temporal features of wind speed dynamics and integrates the Wasserstein distance into the discriminator to enhance the stability and robustness of the imputation process. The framework was tested using in-situ wind speed data from four meteorological stations operated by the Hong Kong Observatory. Comprehensive evaluations were conducted in both single-station and multi-station imputation scenarios. The experimental results demonstrate the exceptional performance of the proposed method, with accurate recovery of missing data even under challenging conditions, including scenarios with a high missing rate of up to 80%.

No abstract

The Global Historical Climatology Network-Daily database contains, among other variables, daily maximum and minimum temperatures from weather stations around the globe. It is long known that climatological summary statistics based on daily temperature minima and maxima will not be accurate, if the bias due to the time at which the observations were collected is not accounted for. Despite some previous work, to our knowledge, there does not exist a satisfactory solution to this important problem. In this paper, we carefully detail the problem and develop a novel approach to address it. Our idea is to impute the hourly temperatures at the location of the measurements by borrowing information from the nearby stations that record hourly temperatures, which then can be used to create accurate summaries of temperature extremes. The key difficulty is that these imputations of the temperature curves must satisfy the constraint of falling between the observed daily minima and maxima, and attaining those values at least once in a twenty-four hour period. We develop a spatiotemporal Gaussian process model for imputing the hourly measurements from the nearby stations, and then develop a novel and easy to implement Markov Chain Monte Carlo technique to sample from the posterior distribution satisfying the above constraints. We validate our imputation model using hourly temperature data from four meteorological stations in Iowa, of which one is hidden and the data replaced with daily minima and maxima, and show that the imputed temperatures recover the hidden temperatures well. We also demonstrate that our model can exploit information contained in the data to infer the time of daily measurements.

Abstract The WIVERN (WInd VElocity Radar Nephoscope) mission, currently under the Phase‐0 of the ESA Earth Explorer program, promises to provide new insight in the coupling between winds and microphysics by globally observing, for the first time, vertical profiles of horizontal winds in cloudy areas. The objective of this work is to explore the potential of the WIVERN conically scanning Doppler 94 GHz radar for filling the wind observation gap inside tropical cyclones (TCs). To this aim, realistic WIVERN notional observations of TCs are produced by combining the CloudSat 94 GHz radar reflectivity observations from 2007 to 2009 with ECMWF co‐located winds. Despite the short wavelength of the radar (3 mm), which causes strong attenuation in presence of large amount of liquid hydrometeors, the system can profile most of the TCs, particularly the cloudy areas above the freezing level and the precipitating stratiform regions. The statistical analysis of the results shows that, (a) because of its lower sensitivity, a nadir pointing WIVERN would detect 75% of the clouds observed by CloudSat (45% of winds with 3 m s −1 accuracy, in comparison to CloudSat sampling of clouds), (b) but thanks to its scanning capability, WIVERN would actually provide 53 times more observations of clouds than CloudSat in TCs (30 times more observations of horizontal winds), (c) this corresponds to about 350 (200) million observations of clouds (accurate winds) every year. Such observations could be used to shed light on the physical processes underpinning the evolution of TCs and in data assimilation in order to improve numerical weather prediction.

Abstract Vertical profiles of atmospheric temperature, moisture, wind, and aerosols are essential information for weather monitoring and prediction. Their availability, however, is limited in space and time due to the significant resources required to observe them. To fill this gap, the New York State Mesonet (NYSM) Profiler Network has been deployed as a national testbed to facilitate the research, development and evaluation of ground-based profiling technologies and applications. The testbed comprises 17 profiler stations across the state, forming a long-term regional observational network. Each Profiler station is comprised of a ground-based Doppler lidar, a microwave radiometer (MWR) and an environmental Sky Imaging Radiometer (eSIR). Thermodynamic profiles (temperature and humidity) from the MWR; wind and aerosol profiles from the Doppler lidar; and solar radiance and optical depth parameters from the eSIR are collected, processed, disseminated, and archived every 10 minutes. This paper introduces the NYSM Profiler Network and reviews the network design and siting, instrumentation, network operations and maintenance, data and products, and some example applications highlighting the benefits of the network. Some sample applications include improved situational awareness and monitoring of the sea/land breeze, long-range wildfire smoke transport, air quality (PM 2.5 and AOD) and boundary layer height. Ground-based profiling systems promise a path forward for filling a critical gap in the nation’s observing system with the potential to improve analysis and prediction for many weather-sensitive sectors, such as aviation, ground transportation, health, and wind energy.

Abstract. In July 2018, the International Society for Atmospheric Research using Remotely piloted Aircraft (ISARRA) hosted a flight week to showcase the role remotely piloted aircraft systems (RPASs) can have in filling the atmospheric data gap. This campaign was called Lower Atmospheric Process Studies at Elevation – a Remotely-piloted Aircraft Team Experiment (LAPSE-RATE). In support of this campaign, ground-based remote and in situ systems were also deployed for the campaign. The University of Oklahoma deployed the Collaborative Lower Atmospheric Mobile Profiling System (CLAMPS), the University of Colorado deployed two Doppler wind lidars, and the National Severe Storms Laboratory deployed a mobile mesonet with the ability to launch radiosondes. This paper focuses on the data products from these instruments that result in profiles of the atmospheric state. The data are publicly available in the Zenodo LAPSE-RATE community portal (https://zenodo.org/communities/lapse-rate/, 19 January 2021). The profile data discussed are available at https://doi.org/10.5281/zenodo.3780623 (Bell and Klein, 2020), https://doi.org/10.5281/zenodo.3780593 (Bell et al., 2020b), https://doi.org/10.5281/zenodo.3727224 (Bell et al., 2020a), https://doi.org/10.5281/zenodo.3738175 (Waugh, 2020b), https://doi.org/10.5281/zenodo.3720444 (Waugh, 2020a), and https://doi.org/10.5281/zenodo.3698228 (Lundquist et al., 2020).

Abstract. Global measurements of horizontal winds in cloud and precipitation systems represent a gap in the global observation system. The Wind Velocity Radar Nephoscope (WIVERN) mission, one of the two candidates to be the ESA's Earth Explorer 11 mission, aims at filling this gap based on a conically scanning W-band Doppler radar instrument. The determination of the antenna boresight mispointing angles and the impact of their uncertainty on the line of sight Doppler velocities is critical to achieve the mission requirements. While substantial industrial efforts are on their way to achieving accurate determination of the pointing, alternative (external) calibration approaches are currently under scrutiny. The correction of the line of sight Doppler velocity error introduced by the mispointing only needs knowledge of such mispointing angles and does not need the correction of the mispointing itself. Thus, this work discusses four methods applicable to the WIVERN radar that can be used at different timescales to characterize the antenna mispointing both in the azimuthal and in the elevation directions and to correct the error in the Doppler velocity induced by such mispointing. Results show that elevation mispointing is well corrected at very short timescales by monitoring the range at which the surface peak occurs. Azimuthal mispointing is harder but can be tackled by using the expected profiles of the non-moving surface Doppler velocity. Biases in pointing at longer timescales can be monitored by using a well-established reference database (e.g. ECMWF reanalysis) or ad-hoc ground-based calibrators. Although tailored to the WIVERN mission, the proposed methodologies can be extended to other Doppler concepts featuring conically scanning or slant viewing Doppler systems.

The resolution of regional numerical weather prediction (NWP) models has continuously been increased over the past decades, in part, thanks to the improved computational capabilities. At such small scales, the fast weather evolution is driven by wind rather than by temperature and pressure. Over the ocean and in the free troposphere, where global NWP models are not able to resolve wind scales below 150 km, regional models provide wind dynamics and variance equivalent to 25 km or lower. However, although this variance is realistic, it often results in spurious circulation (e.g., moist convection systems), thus misleading weather forecasts and interpretation. An accurate and consistent initialization of the evolution of the 3-dimensional (3-D) wind structure is therefore essential in regional weather analysis. The wind profiles provided by the ESA Aeolus satellite mission will help filling the observational gap in the upper air and hopefully improve regional weather forecast. For a correct assimilation into NWP models, the observations need to be characterized in terms of their spatial scales and measurement errors. To this end, the triple collocation method, widely used in scatterometry, is applied to Aeolus observations collocated with Mode-S aircraft observations and ECMWF model output. An algorithm for collocating 4D wind observations from Aeolus, Mode-S and ECMWF over a region of Western Europe will be presented, along with measurement errors obtained from triple collocation.

Abstract We investigate the variations of the ocean CO 2 sink during the past three decades using global surface ocean maps of the partial pressure of CO 2 reconstructed from observations contained in the Surface Ocean CO 2 Atlas Version 2. To create these maps, we used the neural network‐based data interpolation method of Landschützer et al. (2014) but extended the work in time from 1998 to 2011 to the period from 1982 through 2011. Our results suggest strong decadal variations in the global ocean carbon sink around a long‐term increase that corresponds roughly to that expected from the rise in atmospheric CO 2 . The sink is estimated to have weakened during the 1990s toward a minimum uptake of only −0.8 ± 0.5 Pg C yr −1 in 2000 and thereafter to have strengthened considerably to rates of more than −2.0 ± 0.5 Pg C yr −1 . These decadal variations originate mostly from the extratropical oceans, while the tropical regions contribute primarily to interannual variations. Changes in sea surface temperature affecting the solubility of CO 2 explain part of these variations, particularly at subtropical latitudes. But most of the higher‐latitude changes are attributed to modifications in the surface concentration of dissolved inorganic carbon and alkalinity, induced by decadal variations in atmospheric forcing, with patterns that are reminiscent of those of the Northern and Southern Annular Modes. These decadal variations lead to a substantially smaller cumulative anthropogenic CO 2 uptake of the ocean over the 1982 through 2011 period (reduction of 7.5 ± 5.5 Pg C) relative to that derived by the Global Carbon Budget.

Abstract In mountain terrain, well-configured high-resolution atmospheric models are able to simulate total annual rain and snowfall better than spatial estimates derived from in situ observational networks of precipitation gauges, and significantly better than radar or satellite-derived estimates. This conclusion is primarily based on comparisons with streamflow and snow in basins across the western United States and in Iceland, Europe, and Asia. Even though they outperform gridded datasets based on gauge networks, atmospheric models still disagree with each other on annual average precipitation and often disagree more on their representation of individual storms. Research to address these difficulties must make use of a wide range of observations (snow, streamflow, ecology, radar, satellite) and bring together scientists from different disciplines and a wide range of communities.

Abstract Quantifying rates of climate change in mountain regions is of considerable interest, not least because mountains are viewed as climate “hotspots” where change can anticipate or amplify what is occurring elsewhere. Accelerating mountain climate change has extensive environmental impacts, including depletion of snow/ice reserves, critical for the world's water supply. Whilst the concept of elevation‐dependent warming (EDW), whereby warming rates are stratified by elevation, is widely accepted, no consistent EDW profile at the global scale has been identified. Past assessments have also neglected elevation‐dependent changes in precipitation. In this comprehensive analysis, both in situ station temperature and precipitation data from mountain regions, and global gridded data sets (observations, reanalyses, and model hindcasts) are employed to examine the elevation dependency of temperature and precipitation changes since 1900. In situ observations in paired studies (using adjacent stations) show a tendency toward enhanced warming at higher elevations. However, when all mountain/lowland studies are pooled into two groups, no systematic difference in high versus low elevation group warming rates is found. Precipitation changes based on station data are inconsistent with no systematic contrast between mountain and lowland precipitation trends. Gridded data sets (CRU, GISTEMP, GPCC, ERA5, and CMIP5) show increased warming rates at higher elevations in some regions, but on a global scale there is no universal amplification of warming in mountains. Increases in mountain precipitation are weaker than for low elevations worldwide, meaning reduced elevation‐dependency of precipitation, especially in midlatitudes. Agreement on elevation‐dependent changes between gridded data sets is weak for temperature but stronger for precipitation.

With the Arctic rapidly changing, the needs to observe, understand, and model the changes are essential. To support these needs, an annual cycle of observations of atmospheric properties, processes, and interactions were made while drifting with the sea ice across the central Arctic during the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition from October 2019 to September 2020. An international team designed and implemented the comprehensive program to document and characterize all aspects of the Arctic atmospheric system in unprecedented detail, using a variety of approaches, and across multiple scales. These measurements were coordinated with other observational teams to explore cross-cutting and coupled interactions with the Arctic Ocean, sea ice, and ecosystem through a variety of physical and biogeochemical processes. This overview outlines the breadth and complexity of the atmospheric research program, which was organized into 4 subgroups: atmospheric state, clouds and precipitation, gases and aerosols, and energy budgets. Atmospheric variability over the annual cycle revealed important influences from a persistent large-scale winter circulation pattern, leading to some storms with pressure and winds that were outside the interquartile range of past conditions suggested by long-term reanalysis. Similarly, the MOSAiC location was warmer and wetter in summer than the reanalysis climatology, in part due to its close proximity to the sea ice edge. The comprehensiveness of the observational program for characterizing and analyzing atmospheric phenomena is demonstrated via a winter case study examining air mass transitions and a summer case study examining vertical atmospheric evolution. Overall, the MOSAiC atmospheric program successfully met its objectives and was the most comprehensive atmospheric measurement program to date conducted over the Arctic sea ice. The obtained data will support a broad range of coupled-system scientific research and provide an important foundation for advancing multiscale modeling capabilities in the Arctic.

ABSTRACT We created a new dataset of spatially interpolated monthly climate data for global land areas at a very high spatial resolution (approximately 1 km 2 ). We included monthly temperature (minimum, maximum and average), precipitation, solar radiation, vapour pressure and wind speed, aggregated across a target temporal range of 1970–2000, using data from between 9000 and 60 000 weather stations. Weather station data were interpolated using thin‐plate splines with covariates including elevation, distance to the coast and three satellite‐derived covariates: maximum and minimum land surface temperature as well as cloud cover, obtained with the MODIS satellite platform. Interpolation was done for 23 regions of varying size depending on station density. Satellite data improved prediction accuracy for temperature variables 5–15% (0.07–0.17 °C), particularly for areas with a low station density, although prediction error remained high in such regions for all climate variables. Contributions of satellite covariates were mostly negligible for the other variables, although their importance varied by region. In contrast to the common approach to use a single model formulation for the entire world, we constructed the final product by selecting the best performing model for each region and variable. Global cross‐validation correlations were ≥ 0.99 for temperature and humidity, 0.86 for precipitation and 0.76 for wind speed. The fact that most of our climate surface estimates were only marginally improved by use of satellite covariates highlights the importance having a dense, high‐quality network of climate station data.

Abstract. High-spatial-resolution and long-term climate data are highly desirable for understanding climate-related natural processes. China covers a large area with a low density of weather stations in some (e.g., mountainous) regions. This study describes a 0.5′ (∼ 1 km) dataset of monthly air temperatures at 2 m (minimum, maximum, and mean proxy monthly temperatures, TMPs) and precipitation (PRE) for China in the period of 1901–2017. The dataset was spatially downscaled from the 30′ Climatic Research Unit (CRU) time series dataset with the climatology dataset of WorldClim using delta spatial downscaling and evaluated using observations collected in 1951–2016 by 496 weather stations across China. Prior to downscaling, we evaluated the performances of the WorldClim data with different spatial resolutions and the 30′ original CRU dataset using the observations, revealing that their qualities were overall satisfactory. Specifically, WorldClim data exhibited better performance at higher spatial resolution, while the 30′ original CRU dataset had low biases and high performances. Bicubic, bilinear, and nearest-neighbor interpolation methods employed in downscaling processes were compared, and bilinear interpolation was found to exhibit the best performance to generate the downscaled dataset. Compared with the evaluations of the 30′ original CRU dataset, the mean absolute error of the new dataset (i.e., of the 0.5′ dataset downscaled by bilinear interpolation) decreased by 35.4 %–48.7 % for TMPs and by 25.7 % for PRE. The root-mean-square error decreased by 32.4 %–44.9 % for TMPs and by 25.8 % for PRE. The Nash–Sutcliffe efficiency coefficients increased by 9.6 %–13.8 % for TMPs and by 31.6 % for PRE, and correlation coefficients increased by 0.2 %–0.4 % for TMPs and by 5.0 % for PRE. The new dataset could provide detailed climatology data and annual trends of all climatic variables across China, and the results could be evaluated well using observations at the station. Although the new dataset was not evaluated before 1950 owing to data unavailability, the quality of the new dataset in the period of 1901–2017 depended on the quality of the original CRU and WorldClim datasets. Therefore, the new dataset was reliable, as the downscaling procedure further improved the quality and spatial resolution of the CRU dataset and was concluded to be useful for investigations related to climate change across China. The dataset presented in this article has been published in the Network Common Data Form (NetCDF) at https://doi.org/10.5281/zenodo.3114194 for precipitation (Peng, 2019a) and https://doi.org/10.5281/zenodo.3185722 for air temperatures at 2 m (Peng, 2019b) and includes 156 NetCDF files compressed in zip format and one user guidance text file.

Abstract We outline a new and improved uncertainty analysis for the Goddard Institute for Space Studies Surface Temperature product version 4 (GISTEMP v4). Historical spatial variations in surface temperature anomalies are derived from historical weather station data and ocean data from ships, buoys, and other sensors. Uncertainties arise from measurement uncertainty, changes in spatial coverage of the station record, and systematic biases due to technology shifts and land cover changes. Previously published uncertainty estimates for GISTEMP included only the effect of incomplete station coverage. Here, we update this term using currently available spatial distributions of source data, state‐of‐the‐art reanalyses, and incorporate independently derived estimates for ocean data processing, station homogenization, and other structural biases. The resulting 95% uncertainties are near 0.05 °C in the global annual mean for the last 50 years and increase going back further in time reaching 0.15 °C in 1880. In addition, we quantify the benefits and inherent uncertainty due to the GISTEMP interpolation and averaging method. We use the total uncertainties to estimate the probability for each record year in the GISTEMP to actually be the true record year (to that date) and conclude with 87% likelihood that 2016 was indeed the hottest year of the instrumental period (so far).

Weather and climate affect many ecological processes, making spatially continuous yet fine-resolution weather data desirable for ecological research and predictions. Numerous downscaled weather data sets exist, but little attempt has been made to evaluate them systematically. Here we address this shortcoming by focusing on four major questions: (1) How accurate are downscaled, gridded climate data sets in terms of temperature and precipitation estimates? (2) Are there significant regional differences in accuracy among data sets? (3) How accurate are their mean values compared with extremes? (4) Does their accuracy depend on spatial resolution? We compared eight widely used downscaled data sets that provide gridded daily weather data for recent decades across the United States. We found considerable differences among data sets and between downscaled and weather station data. Temperature is represented more accurately than precipitation, and climate averages are more accurate than weather extremes. The data set exhibiting the best agreement with station data varies among ecoregions. Surprisingly, the accuracy of the data sets does not depend on spatial resolution. Although some inherent differences among data sets and weather station data are to be expected, our findings highlight how much different interpolation methods affect downscaled weather data, even for local comparisons with nearby weather stations located inside a grid cell. More broadly, our results highlight the need for careful consideration among different available data sets in terms of which variables they describe best, where they perform best, and their resolution, when selecting a downscaled weather data set for a given ecological application.

Abstract Long‐term time series of key climate variables with a relevant spatiotemporal resolution are essential for environmental science. Moreover, such spatially continuous data, based on weather observations, are commonly used in, e.g., downscaling and bias correcting of climate model simulations. Here we conducted a comprehensive spatial interpolation scheme where seven climate variables (daily mean, maximum, and minimum surface air temperatures, daily precipitation sum, relative humidity, sea level air pressure, and snow depth) were interpolated over Finland at the spatial resolution of 10 × 10 km 2 . More precisely, (1) we produced daily gridded time series ( FMI _ ClimGrid ) of the variables covering the period of 1961–2010, with a special focus on evaluation and permutation‐based uncertainty estimates, and (2) we investigated temporal trends in the climate variables based on the gridded data. National climate station observations were supplemented by records from the surrounding countries, and kriging interpolation was applied to account for topography and water bodies. For daily precipitation sum and snow depth, a two‐stage interpolation with a binary classifier was deployed for an accurate delineation of areas with no precipitation or snow. A robust cross‐validation indicated a good agreement between the observed and interpolated values especially for the temperature variables and air pressure, although the effect of seasons was evident. Permutation‐based analysis suggested increased uncertainty toward northern areas, thus identifying regions with suboptimal station density. Finally, several variables had a statistically significant trend indicating a clear but locally varying signal of climate change during the last five decades.

Reanalysis data are being increasingly used as gridded weather data sources for assessing crop-reference evapotranspiration (ET0) in irrigation water-budget analyses at regional scales. This study assesses the performances of ET0 estimates based on weather data, respectively produced by two high-resolution reanalysis datasets: UERRA MESCAN-SURFEX (UMS) and ERA5-Land (E5L). The study is conducted in Campania Region (Southern Italy), with reference to the irrigation seasons (April–September) of years 2008–2018. Temperature, wind speed, vapor pressure deficit, solar radiation and ET0 derived from reanalysis datasets, were compared with the corresponding estimates obtained by spatially interpolating data observed by a network of 18 automatic weather stations (AWSs). Statistical performances of the spatial interpolations were evaluated with a cross-validation procedure, by recursively applying universal kriging or ordinary kriging to the observed weather data. ERA5-Land outperformed UMS both in weather data and ET0 estimates. Averaging over all 18 AWSs sites in the region, the normalized BIAS (nBIAS) was found less than 5% for all the databases. The normalized RMSE (nRMSE) for ET0 computed with E5L data was 17%, while it was 22% with UMS data. Both performances were not far from those obtained by kriging interpolation, which presented an average nRMSE of 14%. Overall, this study confirms that reanalysis can successfully surrogate the unavailability of observed weather data for the regional assessment of ET0.

Abstract The demand for meteorological gridded datasets has increased within the last few years to inform studies such those in climate, weather, and agriculture. These studies require those data to be readily usable in standard formats with continuous spatial and temporal coverage. Since 2016, Brazil has a daily gridded meteorological data set with spatial resolution of 0.25° × 0.25° from January 1, 1980 to December 31, 2013 which was well received by the community. The main objective of this work is to improve the Brazilian meteorological data set. We do this by increasing the resolution of the minimum and maximum temperature (Tmax and Tmin) gridded interpolations from 0.25° × 0.25° to 0.1° × 0.1° by incorporating data on topographic relief, and increasing the time period covered (January 1, 1961–July 31, 2020). Besides Tmax and Tmin, we also gridded precipitation (pr), solar radiation (Rs), wind speed (u2), and relative humidity (RH) using observed data from 11,473 rain gauges and 1,252 weather stations. By means of ranked cross‐validation statistics, we selected the best interpolation among inverse distance weighting and angular distance weighting methods. We determined that interpolations for Tmax and Tmin are improved by using the elevation of a query point, that accounts for topographic relief, and a temperature lapse rate. Even though this new version has ≈25 years more data relative to the previous one, statistics from cross‐validation were similar. To allow researchers to assess the performance of the interpolation relative to station data in the area, we provide two types of gridded controls.

No abstract

Abstract. The number of personal weather stations (PWSs) with data available through the internet is increasing gradually in many parts of the world. The purpose of this study is to investigate the applicability of these data for the spatial interpolation of precipitation using a novel approach based on indicator correlations and rank statistics. Due to unknown errors and biases of the observations, rainfall amounts from the PWS network are not considered directly. Instead, it is assumed that the temporal order of the ranking of these data is correct. The crucial step is to find the stations which fulfil this condition. This is done in two steps – first, by selecting the locations using the time series of indicators of high precipitation amounts. Then, the remaining stations are then checked for whether they fit into the spatial pattern of the other stations. Thus, it is assumed that the quantiles of the empirical distribution functions are accurate. These quantiles are then transformed to precipitation amounts by a quantile mapping using the distribution functions which were interpolated from the information from the German National Weather Service (Deutscher Wetterdienst – DWD) data only. The suggested procedure was tested for the state of Baden-Württemberg in Germany. A detailed cross validation of the interpolation was carried out for aggregated precipitation amount of 1, 3, 6, 12 and 24 h. For each of these temporal aggregations, nearly 200 intense events were evaluated, and the improvement of the interpolation was quantified. The results show that the filtering of observations from PWSs is necessary as the interpolation error after the filtering and data transformation decreases significantly. The biggest improvement is achieved for the shortest temporal aggregations.

No abstract

No abstract

Abstract. High-spatial-resolution and long-term climate data are highly desirable for understanding climate-related natural processes. China covers a large area with a low density of weather stations in some regions, especially in mountainous regions. This study describes a 0.5' (~ 1 km) dataset of monthly air temperatures at 2 m (minimum, maximum, and mean TMPs) and precipitation (PRE) for China from 1901–2017. The dataset was spatially downscaled from 30' climatic research unit (CRU) time series dataset with the climatology dataset of WorldClim by using Delta spatial downscaling and evaluated using observations during 1951–2016 from 496 weather stations across China. Moreover, the bicubic, bilinear, and nearest-neighbor interpolation methods were compared in the downscaling processes. Among the three interpolation methods, bilinear interpolation exhibited the best performance to generate the downscaled dataset. Compared with the evaluations of the original CRU dataset, the mean absolute error of the new dataset (i.e., 0.5' downscaled dataset with the bilinear interpolation) relatively decreased by 35.4 %–48.7 % for TMPs and 25.7 % for PRE, the root-mean-square error relatively decreased by 32.4 %–44.9 % for TMPs and 25.8 % for PRE, the Nash–Sutcliffe efficiency coefficients relatively increased by 9.6 %–13.8 % for TMPs and 31.6 % for PRE, and the correlation coefficients relatively increased by 0.2 %–0.4 % for TMPs and 5.0 % for PRE. Further, the new dataset could provide detailed climatology data and annual trend of each climatic variable across China, and the results could be well evaluated using observations at the station. Although the evaluation of new dataset was not carried out before 1950 owing to a lack of data availability, the downscaling procedure used data from CRU and WordClim and did not incorporate observations. Thus the quality of the new dataset before 1950 mainly depended on that of the CRU and WordClim datasets. The evaluations showed that the overall quality of the CRU and WordClim datasets was satisfactory, and the downscaling procedure further improved the quality and spatial resolution of the CRU dataset. The new dataset will be useful in investigations related to climate change across China. The dataset presented in this article has been published in Network Common Data Form (NetCDF) at http://doi.org/10.5281/zenodo.3114194 for precipitation (Peng, 2019a) and http://doi.org/10.5281/zenodo.3185722 for air temperatures at 2 m (Peng, 2019b). The dataset includes 156 NetCDF files compressed with zip format and one user guidance text file.

No abstract

Abstract. An accurate spatially continuous air temperature data set is crucial for multiple applications in the environmental and ecological sciences. Existing spatial interpolation methods have relatively low accuracy, and the resolution of available long-term gridded products of air temperature for China is coarse. Point observations from meteorological stations can provide long-term air temperature data series but cannot represent spatially continuous information. Here, we devised a method for spatial interpolation of air temperature data from meteorological stations based on powerful machine learning tools. First, to determine the optimal method for interpolation of air temperature data, we employed three machine learning models: random forest, support vector machine, and Gaussian process regression. A comparison of the mean absolute error, root mean square error, coefficient of determination, and residuals revealed that a Gaussian process regression had high accuracy and clearly outperformed the other two models regarding the interpolation of monthly maximum, minimum, and mean air temperatures. The machine learning methods were compared with three traditional methods used frequently for spatial interpolation: inverse distance weighting, ordinary kriging, and ANUSPLIN (Australian National University Spline). Results showed that the Gaussian process regression model had higher accuracy and greater robustness than the traditional methods regarding interpolation of monthly maximum, minimum, and mean air temperatures in each month. A comparison with the TerraClimate (Monthly Climate and Climatic Water Balance for Global Terrestrial Surfaces), FLDAS (Famine Early Warning Systems Network (FEWS NET) Land Data Assimilation System), and ERA5 (ECMWF, European Centre for Medium-Range Weather Forecasts, Climate Reanalysis) data sets revealed that the accuracy of the temperature data generated using the Gaussian process regression model was higher. Finally, using the Gaussian process regression method, we produced a long-term (January 1951 to December 2020) gridded monthly air temperature data set, with 1 km resolution and high accuracy for China, which we named GPRChinaTemp1km. The data set consists of three variables: monthly mean air temperature, monthly maximum air temperature, and monthly minimum air temperature. The obtained GPRChinaTemp1km data were used to analyse the spatiotemporal variations of air temperature using Theil–Sen median trend analysis in combination with the Mann–Kendall test. It was found that the monthly mean and minimum air temperatures across China were characterised by a significant trend of increase in each month, whereas monthly maximum air temperatures showed a more spatially heterogeneous pattern, with significant increase, non-significant increase, and non-significant decrease. The GPRChinaTemp1km data set is publicly available at https://doi.org/10.5281/zenodo.5112122 (He et al., 2021a) for monthly maximum air temperature, at https://doi.org/10.5281/zenodo.5111989 (He et al., 2021b) for monthly mean air temperature, and at https://doi.org/10.5281/zenodo.5112232 (He et al., 2021c) for monthly minimum air temperature.

Measuring the amount of rainfall is essential for a wide-area evaluation of the risk of landslide disaster using a real-time simulation. In Thailand, located in Monsoon Asia, point observation is conducted using a rain gauge. Interpolation calculation is crucial for obtaining the planar rainfall intensity for the wide-area analysis from scattered point observation data. In this study, to accurately calculate rainfall intensity using the inverse distance weighting (IDW) method, the parameters affecting the results are examined. Additionally, using obtained rainfall data, a simple prediction calculation of groundwater level fluctuation by Wakai et al. [1] and Ozaki et al. [2] is performed. Finally, the relationship between the rainfall intensity and the fluctuation of groundwater level will be discussed.

Abstract The Canadian forest fire weather index (FWI) system requires spatially continuous, gridded weather data for temperature, relative humidity, wind speed, and precipitation. Reliable estimates of the Canadian FWI system components are needed to ensure the safety of communities, resources, and ecosystems. The quality of the interpolated input weather variables are typically evaluated using error estimates from cross‐validation. These error estimates are used for selecting between spatial interpolation methods for generating the continuous weather surfaces. Leave‐one‐out cross‐validation (LOOCV) is the most commonly used method, but it is biased in spatially clustered weather station networks. Accurate error estimation is important for selecting the optimal interpolation method and evaluating how well an interpolated surface represents true patterns in a weather variable. Other cross‐validation methods may better account for bias relating to clustered weather station networks. We present a comparison of cross‐validation methods for evaluating spatial interpolation models of weather variables for generating the inputs to the Canadian FWI system with the objective of determining whether they identify the same spatial interpolation model as having the lowest error. We found that LOOCV, shuffle‐split, stratified shuffle‐split, and a modified buffered leave‐one‐out procedure generally identified the same spatial interpolation models as having the lowest error. Spatial k‐ fold favored spatial interpolation models with extrapolation ability. Our findings indicate that the most computationally efficient cross‐validation approach can be used for automatically selecting spatial interpolation models for weather surface generation, which will improve the quality of historical daily FWI maps.

Abstract Spatially continuous data products are essential for a number of applications including climate and hydrologic modeling, weather prediction, and water resource management. In this work, a distance-weighted interpolation method used to map daily rainfall and temperature in Hawaii is described and assessed. New high-resolution (250 m) maps were developed for daily rainfall and daily maximum ( T max ) and minimum ( T min ) near-surface air temperature for the period 1990–2014. Maps were produced using climatologically aided interpolation, in which station anomalies were interpolated using an optimized inverse distance weighting approach and then combined with long-term means to produce daily gridded estimates. Leave-one-out cross validation was performed to assess the quality of the final daily grids. The median absolute prediction error for rainfall was 0.1 mm with an average overprediction (+0.6 mm) on days when total rainfall was less than 1 mm. On days with total rainfall greater than 1 mm, median absolute prediction errors were 2 mm and rainfall was typically underpredicted above the 10-mm threshold. For daily temperature, median absolute prediction errors were 3.1° and 2.8°C for T max and T min , respectively. On average, this method overpredicted T max (+1.1°C) and T min (+1.5°C), and errors varied considerably among stations. Errors for all variables exhibited significant seasonal variations. However, the annual range of errors was small. The methods presented here provide an effective approach for mapping daily weather fields in a topographically diverse region and improve on previous products in their spatial resolution, time period of coverage, and use of data.

Spatial interpolation of fire weather variables from station data allow fire danger indices to be mapped continuously across the landscape. This information is crucial to fire management agencies, particularly in areas where weather data are sparse. We compare the performance of several standard interpolation methods (inverse distance weighting, spline, and geostatistical interpolation methods) for estimating output from the Canadian Fire Weather Index (FWI) system at unmonitored locations. We find that geostatistical methods (kriging) generally outperform the other methods, particularly when elevation is used as a covariate. We also find that interpolation of the input meteorological variables and the previous day’s moisture codes to unmonitored locations followed by calculation of the FWI output variables is preferable to first calculating the FWI output variables and then interpolating, in contrast to previous studies. Alternatively, when the previous day’s moisture codes are estimated from interpolated weather, rather than directly interpolated, errors can accumulate and become large. This effect is particularly evident for the duff moisture code and drought moisture code due to their significant autocorrelation.

Weather index insurance for crops is at the developmental stage, however, this type of insurance is particularly susceptible to the problem of spatial basis risk. Spatial basis risk occurs when the weather observed at weather stations does not match the weather experienced on the farmer’s property, causing improper indemnities to be paid to the farmer. However, spatial basis risk may be reduced through the use of averaging and spatial interpolation techniques, such as inverse distance weighting and kriging. These techniques make it possible to incorporate multiple weather stations in the estimation process rather than using only the single closest station, potentially resulting in more accurate estimations and thereby reducing spatial basis risk. Therefore, the objective of this study is to examine the extent to which the choice of spatial interpolation techniques can influence the amount of spatial basis risk in a weather-based insurance model. Using forage crops from the province of Ontario, Canada, as an example, a weather insurance index is developed based on cooling degree days. The weather index represents the heat stress that the crops receive over the growing season. This insurance index is used to determine to what extent spatial basis risk can be reduced by the insurer’s choice of spatial interpolation technique. Seven different interpolation methods are applied to temperature data from Ontario, and theoretical indemnities are calculated for forage producers across the province. By analyzing the correlation between the estimated indemnities and reported forage yields, the amount of spatial basis risk in each model is quantified. The results of this study highlight the importance of choosing an appropriate method based on the characteristics of the target region (and data). Operationally this is important because insurers typically apply the same interpolation methods across an entire region. While one finding of this research may suggest that governments and/or insurance companies may wish to invest in additional weather stations to improve the accuracy of the interpolation method and index, this may not be feasible in practice. Given this, future research may consider utilizing satellite-based remote sensing weather estimates to augment the weather station data and reduce basis risk.

Temperature is a driving climate variable for grapevine development and grape ripening kinetics. The current study first reports interpolation of daily minimum and maximum temperature data by a weather station network from 2001 to 2005 in the Bordeaux (France) region by means of regression kriging using terrain, satellite and land-cover derived covariates. Second it analyses the interpolation procedure errors in agroclimatic indices by means of cross validation and then it compares the field observations of grapevine phenology to temperature-based predicted phenology applied to interpolated data. Finally it proposes a simple method to perform a zoning of Bordeaux vineyards based upon the spatialized prediction of the day on which grape sugar content reaches 200 g.L-1. The zoning performed shows large potential differences in grape maturity date (up to 20 days) induced by temperature spatial variability in a low relief area.

Abstract. Snow models are usually evaluated at sites providing high-quality meteorological data, so that the uncertainty in the meteorological input data can be neglected when assessing model performances. However, high-quality input data are rarely available in mountain areas and, in practical applications, the meteorological forcing used to drive snow models is typically derived from spatial interpolation of the available in situ data or from reanalyses, whose accuracy can be considerably lower. In order to fully characterize the performances of a snow model, the model sensitivity to errors in the input data should be quantified. In this study we test the ability of six snow models to reproduce snow water equivalent, snow density and snow depth when they are forced by meteorological input data with gradually lower accuracy. The SNOWPACK, GEOTOP, HTESSEL, UTOPIA, SMASH and S3M snow models are forced, first, with high-quality measurements performed at the experimental site of Torgnon, located at 2160 m a.s.l. in the Italian Alps (control run). Then, the models are forced by data at gradually lower temporal and/or spatial resolution, obtained by (i) sampling the original Torgnon 30 min time series at 3, 6, and 12 h, (ii) spatially interpolating neighbouring in situ station measurements and (iii) extracting information from GLDAS, ERA5 and ERA-Interim reanalyses at the grid point closest to the Torgnon site. Since the selected models are characterized by different degrees of complexity, from highly sophisticated multi-layer snow models to simple, empirical, single-layer snow schemes, we also discuss the results of these experiments in relation to the model complexity. The results show that, when forced by accurate 30 min resolution weather station data, the single-layer, intermediate-complexity snow models HTESSEL and UTOPIA provide similar skills to the more sophisticated multi-layer model SNOWPACK, and these three models show better agreement with observations and more robust performances over different seasons compared to the lower-complexity models SMASH and S3M. All models forced by 3-hourly data provide similar skills to the control run, while the use of 6- and 12-hourly temporal resolution forcings may lead to a reduction in model performances if the incoming shortwave radiation is not properly represented. The SMASH model generally shows low sensitivity to the temporal degradation of the input data. Spatially interpolated data from neighbouring stations and reanalyses are found to be adequate forcings, provided that temperature and precipitation variables are not affected by large biases over the considered period. However, a simple bias-adjustment technique applied to ERA-Interim temperatures allowed all models to achieve similar performances to the control run. Regardless of their complexity, all models show weaknesses in the representation of the snow density.

Abstract Strong historical and predicted future warming over high-latitudes prompt significant effects on agricultural and forest ecosystems. Thus, there is an urgent need for spatially-detailed information of current thermal growing season (GS) conditions and their past changes. Here, we deployed a large network of weather stations, high-resolution geospatial environmental data and semi-parametric regression to model the spatial variation in multiple GS variables (i.e. beginning, end, length, degree day sum [GDDS, base temperature + 5 °C]) and their intra-annual variability and temporal trends in respect to geographical location, topography, water and forest cover, and urban land use variables over northern Europe. Our analyses revealed substantial spatial variability in average GS conditions (1990–2019) and consistent temporal trends (1950–2019). We showed that there have been significant changes in thermal GS towards earlier beginnings (on average 15 days over the study period), increased length (23 days) and GDDS (287 °C days). By using a spatial interpolation of weather station data to a regular grid we predicted current GS conditions at high resolution (100 m × 100 m) and with high accuracy (correlation ≥ 0.92 between observed and predicted mean GS values), whereas spatial variation in temporal trends and interannual variability were more demanding to predict. The spatial variation in GS variables was mostly driven by latitudinal and elevational gradients, albeit they were constrained by local scale variables. The proximity of sea and lakes, and high forest cover suppressed temporal trends and inter-annual variability potentially indicating local climate buffering. The produced high-resolution datasets showcased the diversity in thermal GS conditions and impacts of climate change over northern Europe. They are valuable in various forest management and ecosystem applications, and in adaptation to climate change.

Windstorms result in significant damage and economic loss and are a major recurring threat in many countries. Estimating surface-level wind speeds resulting from windstorms is a complicated problem, but geostatistical spatial interpolation methods present a potential solution. Maximum sustained and peak gust weather station data from two historic windstorms in Europe were analyzed to predict surface-level wind speed surfaces across a large and topographically varied landscape. Disjunctively sampled maximum sustained wind speeds were adjusted to represent equivalent continuously sampled 10-minute wind speeds and missing peak gust station data were estimated by applying a gust factor to the recorded maximum sustained wind speeds. Wind surfaces were estimated based on anisotropic and isotropic kriging interpolation methodologies. The study found that anisotropic kriging is well-suited for interpolating wind speeds in meso- and macro-scale areas because it accounts for wind direction and trends in wind speeds across a large, heterogeneous surface, and resulted in interpolation surface improvement in most models evaluated. Statistical testing of interpolation error for stations stratified by geographic classification revealed that stations in coastal and/or mountainous locations had significantly higher prediction errors when compared with stations in non-coastal/non-mountainous locations. These results may assist in mitigating losses to structures due to excessive wind events.

No abstract

Abstract Extreme heat events cause periodic damage to crop yields and may pose a threat to the income of farmers. Weather index insurance provides payouts to farmers in the case of measurable weather extremes to keep production going. However, its viability depends crucially on the accuracy of local weather indices to predict yield damages from adverse weather conditions. So far, extreme heat indices are poorly represented in weather index insurance. In this study, we construct indices of extreme heat using observations at the nearest weather station and estimates for each county using three interpolation techniques: inverse-distance weighting, ordinary kriging, and regression kriging. Applying these indices to insurance against heat damage to corn in Illinois and Iowa, we show that heat index insurance reduces relative risk premiums by 27%–29% and that interpolated indices outperform the nearest-neighbor index by around 2%–3% in terms of relative risk reduction. Further, we find that the advantage of interpolation over a nearest-neighbor index in terms of relative risk reduction increases as the sample of weather stations is reduced. These findings suggest that heat index insurance can work even when weather data are spatially sparse, which delivers important implications for insurance practice and policy makers. Further, our public code repository provides a rich toolbox of methods to be used for other perils, crops, and regions. Our results are therefore not only replicable but also constitute a cornerstone for projects to come.

Abstract Reproducing realistic date‐ and site‐specific unsteady wind conditions in large‐eddy simulations is becoming increasingly useful in wind energy. How to run a large‐eddy simulation to match observed conditions, however, remains an open research question. One approach that has received considerable attention is mesoscale‐to‐microscale coupling, in which information about the mesoscale weather, most commonly acquired from a mesoscale numerical weather model, is passed on to a microscale model. In this paper, we demonstrate how the recently developed profile‐assimilation technique, a form of mesoscale‐to‐microscale coupling, can be used to drive large‐eddy simulations solely based on observed mean‐flow profiles at a single location, bypassing the need for auxiliary mesoscale simulations. The new approach is evaluated for a diurnal cycle at the Scaled Wind Farm Technology site. Observed mean‐flow profiles from the ground up to a height of 2 km are reconstructed by aggregating measurements from multiple instruments, and gaps in the data are infilled with natural neighbor interpolation. We perform nine simulations using various forcing approaches to deal with data limitations. The results show that it is indeed possible to drive microscale large‐eddy simulation with observations using the profile‐assimilation technique, notwithstanding large gaps in virtual potential temperature measurements. However, profile assimilation with vertical smoothing of the error between the desired and actual profiles is required. Without that smoothing, the microscale simulations develop unrealistically high turbulence levels under many situations. Finally, we show that simulated mesoscale data can account for missing observations, although care is needed as both data sources are not necessarily compatible.

Abstract. The WISSDOM (Wind Synthesis System using Doppler Measurements) synthesis scheme was developed to derive high-resolution 3-dimensional (3D) winds under clear-air conditions. From this variational-based scheme, detailed wind information was obtained from scanning Doppler lidars, automatic weather stations (AWSs), sounding observations, and local reanalysis datasets (LDAPS, Local Data Assimilation and Prediction System), which were utilized as constraints to minimize the cost function. The objective of this study is to evaluate the performance and accuracy of derived 3D winds from this modified scheme. A strong wind event was selected to demonstrate its performance over complex terrain in Pyeongchang, South Korea. The size of the test domain is 12×12 km2 extended up to 3 km a.m.s.l. (above mean sea level) height with a remarkably high horizontal and vertical resolution of 50 m. The derived winds reveal that reasonable patterns were explored from a control run, as they have significant similarity with the sounding observations. The results of intercomparisons show that the correlation coefficients between derived horizontal winds and sounding observations are 0.97 and 0.87 for u- and v-component winds, respectively, and the averaged bias (root mean square deviation, RMSD) of horizontal winds is between −0.78 and 0.09 (1.77 and 1.65) m s−1. The correlation coefficients between WISSDOM-derived winds and lidar QVP (quasi-vertical profile) are 0.84 and 0.35 for u- and v-component winds, respectively, and the averaged bias (RMSD) of horizontal winds is between 2.83 and 2.26 (3.69 and 2.92) m s−1. The statistical errors also reveal a satisfying performance of the retrieved 3D winds; the median values of wind directions are −5 to 5 (0 to 2.5)∘, the wind speed is approximately −1 to 3 m s−1 (−1 to 0.5 m s−1), and the vertical velocity is −0.2 to 0.6 m s−1 compared with the lidar QVP (sounding observations). A series of sensitivity tests with different weighting coefficients, radius of influence (RI) in interpolation, and various combination of different datasets were also performed. The results indicate that the present setting of the control run is the optimal reference to WISSDOM synthesis in this event and will help verify the impacts against various scenarios and observational references in this area.

Mesoscale numerical weather prediction models usually provide information regarding environmental parameters near urban areas at a spatial resolution of the order of thousands or hundreds of meters, at best. If detailed information is required at the building scale, an urban-scale model is necessary. Proper definition of the boundary conditions for the urban-scale simulation is very demanding in terms of its compatibility with environmental conditions and numerical modeling. Here, steady-state computational fluid dynamics (CFD) microscale simulations of the wind and thermal environment are performed over an urban area of Kozani, Greece, using both the k-ε and k-ω SST turbulence models. For the boundary conditions, instead of interpolating vertical profiles from the mesoscale solution, which is obtained with the atmospheric pollution model (TAPM), a novel approach is proposed, relying on previously developed analytic expressions, based on the Monin Obuhkov similarity theory, and one-way coupling with minimal information from mesoscale indices (Vy = 10 m, Ty = 100 m, L*). The extra computational cost is negligible compared to direct interpolation from mesoscale data, and the methodology provides design phase flexibility, allowing for the representation of discrete urban-scale atmospheric conditions, as defined by the mesoscale indices. The results compared favorably with the common interpolation practice and with the following measurements obtained for the current study: SODAR for vertical profiles of wind speed and a meteorological temperature profiler for temperature. The significance of including the effects of diverse atmospheric conditions is manifested in the microscale simulations, through significant variations (~30%) in the critical building-related design parameters, such as the surface pressure distributions and local wind patterns.

Abstract This study proposes an empirical model for preliminary wind‐resist design of downburst flow. Existing empirical models were compared with field data and found to underpredict horizontal wind speed below the height corresponding to the maximum radial velocity, due to the neglect of viscous effects and the evolution of vertical wind profiles along radial direction. To address these deficiencies, semi‐empirical piecewise functions including wall shear effects in the local turbulent boundary layer and interpolation functions were proposed to improve the accuracy of existing models. The wind profile based on Coles' theory was found to agree well with field data, with the parabola interpolation function being the most desirable. Using the proposed method, the vertical profile of horizontal wind speed at different local radial locations can be predicted for wind resist design given the inlet wind speed of the downburst flow. Overall, this model improves upon existing empirical models and allows for more accurate wind‐resist design.

Abstract We describe the construction of a new version of the Europe‐wide E‐OBS temperature (daily minimum, mean, and maximum values) and precipitation data set. This version provides an improved estimation of interpolation uncertainty through the calculation of a 100‐member ensemble of realizations of each daily field. The data set covers the period back to 1950 and provides gridded fields at a spacing of 0.25 ∘ × 0.25 ∘ in regular latitude/longitude coordinates. As with the original E‐OBS data set, the ensemble version is based on the station series collated as part of the ECA&D initiative. Station density varies significantly over the domain, and over time, and a reliable estimation of interpolation uncertainty in the gridded fields is therefore important for users of the data set. The uncertainty quantified by the ensemble data set is more realistic than the uncertainty estimates in the original version, although uncertainty is still underestimated in data‐sparse regions. The new data set is compared against the earlier version of E‐OBS and against regional gridded data sets produced by a selection of National Meteorological Services. In terms of both climatological averages and extreme values, the new version of E‐OBS is broadly comparable to the earlier version. Nonetheless, users will notice differences between the two E‐OBS versions, especially for precipitation, which arises from the different gridding method used.

Abstract This study introduces an efficient and precise method for gathering atmospheric wind field data in specific regions, utilizing unmanned aerial vehicles (UAVs) equipped with anemometers. We conducted outdoor wind field measurements over complex terrains using a six-rotor UAV equipped with an ultrasonic anemometer. The results obtained include a vertical wind field profile at the center of the measured site, along with two planar wind field profiles at 20 m and 40 m heights. The analysis reveals a remarkably high fitting accuracy for the wind profile and turbulence intensity results obtained. Furthermore, the planar wind field data partially demonstrates the impact of terrain on the upper-level wind field within the surveyed area. Lastly, we established a three-dimensional wind field visualization approach using the data gathered through the UAV wind measurement method, implementing Kriging interpolation. This study’s outcomes offer novel insights and methodologies for local wind field measurement, micro-siting in wind farms, and the creation of visualized wind fields.

The Hydrological Cycle in Mediterranean Experiment (HyMeX) is a 10-year international programme devoted to improving our understanding of the hydrological cycle in the Mediterranean area, with special emphasis on the predictability and evolution of high-impact weather events (Drobinski et al., 2014). Heavy precipitation is a major natural hazard in the Mediterranean. Daily surface rainfall greater than 100 mm is not uncommon for Mediterranean precipitation events, and often such amounts are recorded in only a few hours, associated with mesoscale convective systems (MCSs). The occurrence of these heavy precipitation amounts over small river catchments that are characteristic of the Mediterranean region often leads to devastating flash floods and flooding events. Each year, these heavy precipitation events (HPEs) result in up to hundreds of millions of euros in damages and many casualties. The distinctive topography and geographical location of the Mediterranean basin (Figure 1) make the region particularly prone to HPEs. Most of these events occur in autumn over the western Mediterranean when sea water is warmest and serves as an important heat and moisture source from which convective and baroclinic atmospheric systems can derive their energy. The steep orography surrounding the Mediterranean Sea aids in lifting the low-level, conditionally unstable air, thus initiating condensation and convection processes. The HyMeX observation strategy relies on (i) heavily instrumented Special Observation Periods (SOPs) of a few months to provide detailed and specific observations to study key processes, and (ii) longer observation periods repetitively or routinely collecting observations to monitor long-term water cycle processes and rare events such as flash flooding over a few specific instrumented watersheds. The first SOP was dedicated to heavy precipitation and flash floods. This major international field campaign took place from 5 September to 6 November 2012 over the northwestern Mediterranean Sea and its surrounding coastal regions in France, Italy and Spain. Ducrocq et al. (2014) gave a comprehensive description of the SOP1 observation strategy and execution, as well as the detailed list of deployed instruments. The observations collected by more than 200 research instruments constitute an unprecedented dataset. Uniquely in comparison to previous field experiments in the region, the atmosphere, ocean, and land surfaces were all sampled in order to study the interaction and feedbacks between the different components of the hydrological cycle during heavy precipitation and flash floods. As a substantial improvement to previous field campaigns dedicated to heavy precipitation (e.g. MAP field experiment in 1999: Bougeault et al., 2001), SOP1 collected observations also over the sea and of precipitating systems forming over the sea and affecting the coastal areas. A strong modelling component (ocean–atmosphere–hydrology, process–weather, prediction–climate models) was conceived from the beginning in HyMeX, which allowed use of the field campaign data for model validation, data assimilation, improving physical parametrizations and for advancing process understanding. This Special Issue presents a wide range of studies carried out over the last three years exploiting the exceptional dataset of observations and model output collected during SOP1. The special issue consists of a series of 31 articles. Some major results of these articles are summarized in the following sections. Section 2 presents advances in observations and their use in characterizing HPE, section 3 highlights the advances in process understanding, and section 4 details results about predictability of HPE from numerical weather prediction (NWP) and regional climate models. As a primary stage of analysis, the consistency between co-located observations has been assessed for different research instruments measuring winds and water vapour during SOP1 (Chazette et al. (this SI, pp 7–22), Saïd et al. (this SI, pp 23–42)) and satellite products (Rysman et al. (this SI, pp 43–55)). Chazette et al. (this SI, pp 7–22) compare water vapour observations from a ground-based water-vapour Raman lidar (Chazette et al., 2014), an airborne water-vapour lidar and boundary-layer pressurized balloons (Doerenbecher et al., 2016). Saïd et al. (this SI, pp 23–42) compare winds from 11 UHF and VHF wind profiler radars, balloon radiosoundings, in situ aircraft or boundary-layer pressurised balloons. Airborne cloud radar measurements are used to validate the Convective Overshooting (COV) diagnostic based on passive microwave measurements from the Microwave Humidity Sounder (MHS) in Rysman et al. (this SI, pp 43–55). Several studies take advantage of additional measurements from research instruments or availability of data from non-real-time networks to evaluate the performance and adequacy of current operational networks for describing the ambient low-level circulation and atmospheric water vapour. Bock et al. (this SI, pp 56–71) reprocessed more than 1000 ground-based Global Positioning System (GPS) receivers located in Spain, France and Italy to evaluate the accuracy of near-real-time E-GVAP GPS Zenith Total Delay (ZTD) data assimilated in operational NWP systems. They conclude that the mean differences between E-GVAP and reprocessed ZTD data are not negligible. On the other hand, the comparison of the integrated water vapour (IWV) from reprocessed GPS and radiosondes reveals no significant biases over night-time and small biases during daytime. This result gives high confidence in the quality of modern radiosonde systems in contrast to past experiments (e.g. AMMA: Agustí-Panareda et al., 2009). Saïd et al. (this SI, pp 23–42) examine how the wind field derived from UHF and VHF wind profiler radars deployed along the French coast is able to represent the low-level circulation over the northwestern Mediterranean Sea. They clearly show that the drastically different characteristics of the winds induced by the complex terrain of the region make it difficult to retrieve wind fields from the five coastal profiler radars. The study of Khodayar et al. (this SI, pp 72–85) on Intensive Observing Period IOP8 concludes that the variability of water vapour and low-level wind convergence is quite adequately sampled by the operational networks inland, but not over the sea where the spatial and temporal resolution of observations is undeniably insufficient to identify and locate moisture convergence areas. Radar observations collected during SOP1 (Bousquet et al., 2015) represent a valuable dataset that is used to develop and evaluate novel radar-based products for research and operational activities. Bousquet et al. (this SI, pp 86–94) assess the quality of real-time multiple-Doppler radar winds retrieved from the French operational network radars over southern France with the radial velocity measurements collected by an airborne cloud radar during SOP1. The mean difference between the two three-dimensional (3D) cloud wind fields is close to zero and confirms that multiple-Doppler radar wind observations are accurate enough to be used for operational purposes such as NWP model verification. Another radar-based product, namely a hydrometeor classification algorithm, is evaluated in Ribaud et al. (this SI, pp 95–107). They propose an original method to derive 3D hydrometeor fields associated with convective systems from the multi-frequency single-radar hydrometeor classification. Wolfensberger et al. (this SI, pp 108–124) develop a new algorithm to automatically detect the melting layer based on polarimetric radar scans. Raupach and Berne (this SI, pp 125–137) present a new approach for the spatial interpolation of experimental raindrop size distribution (DSD) spectra from a network of disdrometers and a weather radar. Besson et al. (this SI, pp 138–152) show that the refractivity measurements collected from some weather radars during SOP1 compare well with observations from surface weather stations, and are able to capture the diurnal cycle and the low-level conditions during the pre-convection period. The field campaign observations have been extensively used to assess the quality of NWP analyses and forecasts, as well as high-resolution research or regional climate model simulations, before they were exploited to document SOP1 conditions and advance process understanding. In particular, the real-time HyMeX-SOP dedicated version of the convection-permitting AROME NWP system (AROME-WMED: Fourrié et al., 2015) is thoroughly compared to different datasets in Bock et al. (this SI, pp 56–71), Chazette et al. (this SI, pp 7–22), Di Girolamo et al. (this SI, pp 153–172), Duffourg et al. (this SI, pp 259–274), Rainaud et al. (this SI, pp 173–187) and Saïd et al. (this SI, pp 23–42). The characteristics of the IOPs together with the large-scale circulation that prevailed during SOP1 are presented in Ducrocq et al. (2014). IOPs over Spain and Italy are more specifically described in Jansà et al. (2014) and Ferretti et al. (2014), respectively. Monthly precipitation totals from surface stations were well above the corresponding climatology for most regions in October and November 2012, and were near average in September. The overall temporal distribution of HPE is closely related to the Atlantic weather regimes. Rysman et al. (this SI, pp 43–55) shows that convective activity during SOP1, based on the Deep Convection (DC) and COV diagnostics from MHS over the Mediterranean area (including over the sea), was not notably different from the last 12 years, except for some instances in the second half of the SOP period with COV occurrence reaching two to four times the average values. Bock et al. (this SI, pp 56–71) use reprocessed GPS data to examine the spatial and temporal variability of IWV during SOP1. They point out a high temporal variability in the moisture content, with higher content at Mediterranean coastal sites which are under the influence of strong low-level inland advection of moisture. IWV peaks are often observed shortly before precipitation. Saïd et al. (this SI, pp 23–42) highlight the high spatial variability of low-level winds along the northwestern coast, observed by the mesoscale wind profiler radar networks, strongly linked to the complex mountainous coasts and islands and their role in the low-level circulation. The mistral and the tramontane, two regional northerly and northwesterly dry winds which often blow in the northwestern Mediterranean basin, are examples of interactions of the large-scale flow with orography. During autumn, as in SOP1, the second prevailing wind of the northwestern basin is onshore wind, favouring heavy precipitation (Ricard et al., 2012). Di Girolamo et al. (this SI, pp 153–172) study three transition events from mistral/tramontane to southerly marine flow in southern France during SOP1. Low-level wind reversals are found to have a strong impact on water vapour transport, leading to a large variability of the water vapour vertical and horizontal distributions. A noticeable result is that the increase/decrease in water vapour mixing ratio within the boundary layer may be abrupt and marked during these transition periods, with values increasing or decreasing by a factor of 2–4 within 1 h. The high variability of water vapour associated with low-level winds impacts the air–sea fluxes. Rainaud et al. (this SI, pp 173–187) show that the Gulf of Lion is the area with the highest variability of air–sea fluxes during SOP1, due to the prevailing strong dry regional winds (mistral/tramontane). Another remarkable result of this study is that even though some HPEs occur without significant air–sea fluxes, all strong air–sea exchange episodes include, or occur just 1 or 2 days before, HPEs. A major ingredient in HPE is the conditionally unstable, low-level marine flow impinging on the mountainous coastal regions bordering the western Mediterranean Sea, associated with lifting that leads to the onset of deep convection at the same place. The lifting mechanisms mostly result from the interactions of the low-level circulation with the orography, which greatly depends upon the local configuration of the terrain. However, a noteworthy outcome of many of the IOP studies carried out over Spain, France or Italy is that these cases present common characteristics (Figure 1). In addition to the direct role of orographic lifting, they frequently highlight features such as the presence of pre-existing convergence lines (dynamically or orographically induced), the role of a cold pool (possibly, but not necessarily, resulting from evaporative cooling) and contributions of topographical flows (e.g. gap winds and barrier jets). Based on IOP18, IOP19 and pre-SOP1 events over northeastern Italy (NEI), Davolio et al. (this SI, pp 188–205) identify two different dynamical behaviours of the flow impinging on the Alpine orography, associated with two different patterns of heavy precipitation. All the events present a similar initial phase, forced by the advancing synoptic disturbance and characterized by a weak low-level wind coming from the Adriatic Sea (sirocco wind) blocked by the Alps and deflected as an easterly/northeasterly barrier wind over the NEI plain. Then, different interactions with the orography produce two different evolutions: (i) flow-over conditions progressively establish themselves, the barrier wind disappears and the orographically forced uplift of the impinging flow produces intense orographic precipitation over the Alps, with embedded convective activity; (ii) the uplift over the pre-existing cold air over the NEI plain (blocking) is able to initiate convection well upstream of the orography where the unstable incoming flow is forced to rise over the layer characterized by the barrier wind. Blocked flow conditions persist and the convergence line between the sirocco wind and the barrier wind further triggers convection. The precipitation affects the NEI plain or even the coastal area, far from the Alps. Scheffknecht et al. (this SI, pp 206–221) explicitly highlight the role of the deflection of flow around the mountainous islands by studying the localized convective system of IOP15c, which led to severe flooding over southern Corsica. They clearly identify the splitting of the northerly low-level wind around the Corsican orography and the resulting convergence at the southern of the as key mechanisms for this Heavy precipitation only when the between the and western of the flow was and when gap winds from the to et al. (this SI, pp study the role of in initiating convection for two from SOP1. the first the flow marine winds low-level convergence for the convection the flow around the and the resulting convergence is the for the second Another influence of the orography is also on by and Davolio (this SI, pp which the of the and islands as a the rainfall distribution over and Duffourg et al. (this SI, pp study the mesoscale convective systems that over the sea and heavy precipitation over the French coastal region during The convective systems are during their stage over the by and conditionally unstable air carried by a to low-level The low-level wind convergence in this flow is the to initiate and the of convective to the convective The convergence line when a surface in the of the associated with the of an In evaporative low-level convergence and thus This is found to a key role in the deep convection embedded within the associated with the et al. (this SI, pp compare this Mediterranean to deep Mediterranean also associated with heavy precipitation and over the same region and the same of the associated heavy rainfall was by two different (i) deep convection for and (ii) associated with precipitation for the second the most of SOP1, was also characterized by an associated with a and low-level convergence within which an over the coast et al., 2014). et al. (this SI, pp identify two mechanisms leading to heavy precipitation in this (i) a wind convergence line over the sea to the convective and (ii) the interaction of the low-level winds with the orography the heavy precipitation. During SOP1, more than instruments were deployed over southern France to the activity associated with convective systems. the exceptional the 3D of the with a large of recorded during SOP1 et al., Ribaud et al. (this SI, pp the between activity and for the convective line 3D wind and hydrometeor observations from the weather radars. The results that the of from the convective area the area of the is at the of a activity the of from the of the convective area the of the field within this may also have the leading to a On the the of the system over a small may have strongly the resulting in intense activity at high et al. (this SI, pp further study the impact of terrain on precipitation from observations of over southern They show that the orography and the rainfall a role in the rainfall and in the associated processes. The above the of the local other processes, such as or with the near the A low-level flow is a common ingredient for HPEs. The of the moisture over the Mediterranean Sea is quite and and from (i) from the Mediterranean Sea, (ii) from the Atlantic and from Chazette et al. (this SI, pp highlight a of high water-vapour content air associated with the of a and strong low-level winds over the Mediterranean within which during et al. (this SI, pp identify for IOP8 two moisture in the Atlantic for the of the and from the Mediterranean Sea for the et al. (this SI, pp study more specifically the variability of IWV over and particularly the IWV distribution between the and and associated mesoscale processes as well as the strong of IWV in the of the and of the on the upstream coast only small were spatial are more for the of deep convection in the of the than at the to for heavy precipitation have been in years and significant has been the of convection-permitting NWP systems. However, the accuracy of is insufficient to the in of and of rainfall and flash Several studies in this special issue at improving the systems and studying the processes and modelling components that influence The of of new of observations has been by the of which observations from model as a first et al. (this SI, pp develop a weather radar for a wide of such as by with The performance of the with a has been evaluated measurements at and from the French operational weather radar network and also for and The results show a consistency between observations and their but with a in observations by and the by a et al. (this SI, pp study the impact on quality of the of additional SOP1 and satellite In a but small impact is The of the impact depends on the weather that the location of the as well as their with routinely data areas. Scheffknecht et al. (this SI, pp 206–221) study for the to initial conditions and model of convection-permitting They a high to initial of the initial conditions the and location of the On the decreasing the from to not clearly the and Davolio (this SI, pp also a weak impact of a higher horizontal resolution 2 on precipitation or of Rainaud et al. (this SI, pp 173–187) evaluate the low-level atmospheric from over sea the and and measurements during SOP1. A is except during strong mistral/tramontane wind due to (i) a in the and (ii) an of the heat in these cold and dry strong wind by the et al. (this SI, pp the impact of the sea on the low-level fluxes during of the convection-permitting atmospheric the (i) without on the (ii) with and with model output as show that the location of the heavy precipitation on the French is when a is used due to the impact on the resulting in a of the surface wind of the upstream low-level et al. (this SI, pp make use of the SOP1 dataset of HPEs to evaluate the of a operational convection-permitting prediction A result is that a convection-permitting with a few can a with many more data is found not as as a surface be et al. (this SI, pp evaluate and compare precipitation from two convection-permitting systems over the of SOP1. The predictability of heavy precipitation of is found to be strongly related to how the the surface also by Duffourg et al. (this SI, pp and et al. (this SI, pp et al. (this SI, pp study the mechanisms that the predictability over the Atlantic and Mediterranean regions from to the beginning of October during SOP1 when was around the A major outcome about the HPE predictability is that the of interaction between the and the Atlantic the of and thus the conditions over the Mediterranean. forced at the by have been regional climate and a period the SOP1 period within the et al., with the of these regional climate observations and NWP or research models. Khodayar et al. (this SI, pp a comparison of regional climate and convection-permitting NWP on the from the regional climate capture the occurrence of the precipitation they produce notably precipitation totals and than the NWP models. convective events are not well by the regional climate models. The differences with the NWP to in the physical parametrizations variability and vertical distribution of than in the initial or boundary et al. (this SI, pp evaluate two regional climate precipitation datasets over They that of the most intense HPEs were also observed HPEs for most the and they study the between differences and precipitation differences over the years and for SOP1 events. This Special Issue the range of advances have in the understanding and predictability of HPEs. SOP1 a remarkable and dataset that has been to make advances in process understanding, rise to new for heavy precipitation events in the western Mediterranean. has also been in and of NWP and regional climate such as convection-permitting prediction systems or regional models. more studies and are in the coming The of observations from the research instruments deployed during SOP1, such as from the or the airborne is and further advances in understanding and modelling of heavy precipitation events. The role of the Mediterranean Sea and the and of water vapour the precipitating systems be with more integrated to the at The regional climate and more specifically the convection-permitting climate be further of and from Several studies also at improving the key physical parametrizations that strongly influence the of deep convection at and process modelling and modelling in its studies at in the initial by data of new observation or data The of these studies is to and NWP prediction to more of heavy precipitation and of its As for previous major field it take years to the field campaign dataset that long-term the The HyMeX programme is by a large of and are on the HyMeX The field campaign SOP1 was by and The also and the as well as all the that have to the quality of the articles and of this Special

In the last few decades there has been increasing interest in the commercial usage of the stratosphere, especially for Earth observation systems. Stratospheric platforms allow Earth monitoring at a regional scale with persistency toward a limited area. For this reason, accurate meteorological forecasts are needed in order to guarantee stationarity. The main aim of this work is to provide a review of wind prediction techniques in the stratosphere, achieved by the most popular global models, such as ECMWF IFS, NCEP GFS and ICON. Then, the capabilities of the COSMO limited area model to reproduce the wind speed in the stratosphere are evaluated considering a model configuration with very high resolution (about 1 km) over a domain located in Southern Italy, assuming the radio sounding data at Pratica di Mare airport as the reference. Vertical profiles were analyzed for selected days, highlighting good performances, though improvements can be achieved by adopting a fifth-order interpolation of the model data. Finally, monthly wind speed time series for selected heights were post-processed by means of fast Fourier transform, revealing the existence of main frequencies and the presence of a scaling regime and a power law of the form f−β over a broad range of time scales, in the Fourier space. The exponent spectral β is close to the exact 5/3 Kolmogorov value for all the datasets.

Accurate specification of hurricane inner-core structure is critical to predicting the evolution of a hurricane. However, observations over hurricane inner cores are generally lacking. Previous studies have emphasized Tail Doppler radar (TDR) data assimilation to improve hurricane inner-core representation. Recently, Doppler wind lidar (DWL) has been used as an observing system to sample hurricane inner-core and environmental conditions. The NOAA P3 Hurricane Hunter aircraft has DWL installed and can obtain wind data over a hurricane’s inner core when the aircraft passes through the hurricane. In this study, we examine the impact of assimilating DWL winds and TDR radial winds on the prediction of Hurricane Earl (2016) with the NCEP operational Hurricane Weather Research and Forecasting (HWRF) system. A series of data assimilation experiments are conducted with the Gridpoint Statistical Interpolation (GSI)-based ensemble-3DVAR hybrid system to identify the best way to assimilate TDR and DWL data into the HWRF forecast system. The results show a positive impact of DWL data on hurricane analysis and prediction. Compared with the assimilation of u and v components, assimilation of DWL wind speed provides better hurricane track and intensity forecasts. Proper choices of data thinning distances (e.g., 5 km horizontal thinning and 70 hPa vertical thinning for DWL) can help achieve better analysis in terms of hurricane vortex representation and forecasts. In the analysis and forecast cycles, the combined TDR and DWL assimilation (DWL wind speed and TDR radial wind, along with other conventional data, e.g., NCEP Automated Data Processing (ADP) data) offsets the downgrade analysis from the absence of DWL observations in an analysis cycle and outperforms assimilation of a single type of data (either TDR or DWL) and leads to improved forecasts of hurricane track, intensity, and structure. Overall, assimilation of DWL observations has been beneficial for analysis and forecasts in most cases. The outcomes from this study demonstrate the great potential of including DWL wind profiles in the operational HWRF system for hurricane forecast improvement.