Can we measure surface temperature change over Europe and Africa using the EUMETSAT CM SAF Land Surface Temperature CDR?

Abstract

<p>Land surface temperature (LST) measured by satellites is strongly related to the 2m air temperature (T2m) measured in situ. Long-term satellite LST data are therefore an independent measure of surface temperature change in regions where T2m observations are scarce. Recent studies such as Good et al. (2022) suggest that satellite LST can be used to assess warming trends over land if the required homogeneity is assured.</p> <p>The objective of this study is to assess the stability and trends of the upcoming second release of the Satellite Application Facility (SAF) on Climate Monitoring (CM SAF) Land Surface Temperature Climate Data Record (CDR). This unique, hourly-resolved LST CDR spans the entire new WMO climate reference period 1991 to 2020. It is retrieved from the latest version of the well calibrated and quality controlled Meteosat Fundamental Climate Data Record (FCDR) including a recalibration of both the Meteosat SEVIRI and MVIRI thermal channels.  The LST algorithm is based on a radiative transfer-based single channel algorithm to ensure consistency across all Meteosat generations.</p> <p>First, we have assessed the temporal stability of the LST CDR by comparing CM SAF LSTs with T2m air temperatures provided by the newest version the Met Office Hadley Centre global surface temperature dataset (HadCRUT5), homogenized Swiss MetNet stations and ERA5 reanalysis over Africa and Europe. We have assessed the relationship between the LST minus T2m monthly anomalies by calculating the trend and confidence interval of these time series of differences. Observed trends are well below 0.1 K/decade for most stations and regions and are statistically insignificant. This suggests that the CM SAF LST time series are stable.</p> <p>Second, trends in the LST and T2m monthly anomalies are calculated and compared for both clear-sky and all sky T2m for selected regions and stations. LST-based temperature trends in Europe agree well with station data for the period 1991-2020. In Switzerland in-situ based decadal trends are in the range of 0.31K to 0.42K, while satellite-based trends are between 0.28K and 0.39K in flat regions. For mountain stations, we observe a 0.1 K lower trends in satellite-based LSTs compared to in-situ data. This difference is likely caused by difference in spatial sampling between MVIRI and SEVIRI and needs to be further investigated.</p> <p>Overall, this study suggests that the EUMETSAT CM SAF LSTs can be used to augment temperature trends in region with sparse station coverage for the new WMO norm period.  As most climate applications, such as e.g. heat wave monitoring, require stable CDRs to calculate anomaly–based indicators, we also conclude that those stable and independent satellite observations are of high value for climate monitoring. </p> <p>Good, E. J., Aldred, F. M., Ghent, D. J., Veal, K. L., & Jimenez, C. (2022). An analysis of the stability and trends in the LST_cci Land Surface Temperature datasets over Europe. Earth and Space Science, 9, e2022EA002317. https://doi.org/10.1029/2022EA002317</p>