Climatic datasets

We used nine global climatic datasets associated with low-cloud fraction (CF), surface temperature (average, minimum, and maximum) (K), surface pressure (Pa), volumetric solid water content between 7-28 cm below the surface (VSWC) (m3 m-3), precipitation (mm day-1), dew point (K), and potential evapotranspiration (PET) (mm month-1). Except for precipitation, all environmental datasets mentioned were derived directly or indirectly (i.e., PET) from ERA5. ERA5 is a fifth-generation atmospheric reanalysis of global climate produced by the Copernicus Climate Change Service at the European Center for Medium-Range Weather Forecasts (Hersbach et al., 2020). ERA5 is the direct successor to ERA-Interim reanalysis and provides global, hourly historical weather data at a regular grid of 0.25 °(~30 km) (Hersbach et al., 2020). The PET dataset was indirectly derived from ERA5 using the Penman-Monteith equation following the FAO-56 Method (Zotarelli et al., 2010). For this, additional parameters such as wind speed, ground pressure, incoming solar radiation, and clear-sky solar radiation were also obtained from ERA5, as detailed in Table S3 . On the other hand, the precipitation dataset was obtained from CHIRPS (Climate Hazards Infrared Precipitation with Stations) (Funk et al., 2015). The former is a quasi-global rainfall product spanning 50°S – 50°N with 38 years of daily estimations. This precipitation dataset incorporates satellite imagery with in situ station data with a spatial resolution of 0.05° (~ 5.5 km).
Overall, biases around ERA5 and CHIRPS climatic datasets have been evaluated in several studies (Bonshoms et al., 2022; Dommo et al., 2022; McNicholl et al., 2022; Tetzner et al., 2019). Despite this, we acknowledge that the reliability of these global products remains uncertain at varying temporal and spatial scales at certain regions. We used these datasets as state-of-the-art climatic products (Muñoz-Sabater et al., 2021) that may help us to provide a comprehensive overview of the trends that are likely to prevail in these ecosystems and not as factual observations of what these ecosystems are experiencing.