MOTEDAS century: A new high‐resolution secular monthly maximum and minimum temperature grid for the Spanish mainland (1916–2015)

José Carlos Gonzalez‐Hidalgo,Santiago Beguería,Michele Brunetti,Dhais Peña‐Angulo

doi:10.1002/joc.6520

Abstract

AbstractInformation was retrieved from the Annual Climate Summaries (ACS) on average monthly temperatures in Spain for the period 1916–1949, and combined with the digitalized data from the National Climate Data Bank of the Spanish meteorological service (AEMET) to create the new MOTEDAS_century (MOnthly Temperature Dataset of Spain) database of mean monthly maximums (Tmax) and minimums (Tmin). This database was used to calculate a high‐resolution (10 × 10 km) grid that allowed to make a detailed analysis of the temporal evolution of temperature in Spain during the period 1916–2015. Over the period, mean Tmax increased by 1.2°C, while Tmin increased by 1.0°C (this ratio is reversed in the second half of the century). In the last 30 years, however, mean annual Tmax and Tmin trends have not been significant. There are some differences between the new mean annual maximum and minimum temperature series and previous versions, which seem to arise from the much higher number of weather stations analysed, their location and the processing used. There is a discussion on the improved spatial representativeness of the new MOTEDAS_century that provides a more detailed representation of the spatial variability of temperatures in continental Spain.

Highlights

Meteorological records are essential information required in the study of the recent evolution of the climate in the Annual mean value of the monthly means of maximum (Tmax) and temperature series (Tmax) and nocturnal (Tmin) and metadata information of original series can be download at www.Clices.Unizar.es period between the appearance of observation networks in the mid-19th century and remote satellite observations in the 1970s
There is no single procedure or universal algorithm to create the reference series, nor has one of the variables proved to be better than the others, and there is a good amount of. This method could be suitable for creating a grid when there is abundant information, as it is the case of the second half of the 20th century, with the advantage that the number of weather stations used in each time unit is the same and enables uncertainty intervals to be calculated very as has been applied in some previous studies on temperatures in Spain (Gonzalez-Hidalgo et al, 2015)
The series from MOTEDAS (1951–2010) comes from a grid created with 1,358 weather stations homogenized and reconstructed in their missing values, while the AEMET series combines 2,886 high quality-checked stations with over 10 years of data, with the number varying in time (Guijarro, 2013), as with MOTEDAS_century, whose series was calculated with a minimum of 228 stations and, since 1950, always with more than 1,000 observations per month

Summary

| INTRODUCTION

Meteorological records are essential information required in the study of the recent evolution of the climate in the Annual mean value of Tmax and Tmin and metadata information of original series can be download at www.Clices.Unizar.es period between the appearance of observation networks in the mid-19th century and remote satellite observations in the 1970s. There is no single procedure or universal algorithm to create the reference series, nor has one of the variables proved to be better than the others, and there is a good amount of In our opinion, this method could be suitable for creating a grid when there is abundant information, as it is the case of the second half of the 20th century, with the advantage that the number of weather stations used in each time unit (day, month or year) is the same and enables uncertainty intervals to be calculated very as has been applied in some previous studies on temperatures in Spain (Gonzalez-Hidalgo et al, 2015). The trend rate was evaluated by the Theil-Sen slope estimator (Sen, 1968)

| RESULTS

Findings

| DISCUSSION

| CONCLUSIONS