Constructing a meteorological indicator dataset for selected European NUTS 3 regions.

Abstract

The harmonization of data granularity in spatial and temporal terms is an important pre-step to any econometric and machine learning applications. Researchers, who wish to statistically test hypotheses on the relationship between agro-meteorological and European policy outcomes, often observe that agro-meteorological data is typically stored in gridded and temporally detailed form, while many relevant policy outcomes are only available on an aggregated level. This dataset intends to aid empirical investigations by providing a dataset with monthly meteorological indicators on a European Nomenclature of Territorial Units for Statistics level 3 (NUTS 3) regional level for 13 countries for the period from 1989 to 2018. The data we provide allows researchers to investigate hypothesis related to weather volatility and the probability of extreme weather events.We created this dataset from the daily data in grids of 25 km x 25 km provided by the Joint Research Centre of the European Commission. We matched the map with the raw data to a map with the administrative boundaries of European NUTS 3 regions. After appropriately weighting, we calculated the monthly, regional mean, variance and kurtosis of the following variables: maximum, minimum, average air temperature in degrees Centigrade, sum of precipitation in mm and snow depth in cm. We report the covariance between the average temperature and the precipitation as well.