https://crudata.uea.ac.uk/cru/data/drought/ This is the readme file for the self-calibrating Palmer Drought Severity Index (scPDSI) data for the global land surface (excluding Antarctica), spanning the period 1903-2025 with monthly resolution. The spatial resolution is 0.5 x 0.5 degrees. The information below relates to global scPDSI files. The format of the data is a self-describing netCDF file, with information about grid coordinates and time coordinates provided within the file. The grid size is 720 longitudes by 360 latitudes and the coordinates of the CENTRE of the grid cells end in 0.25 or 0.75 (i.e. the EDGES of the grid cells lie on either 0.00 or 0.50 degrees). The grid cells run from 89.75N, 179.75W (centre) to 89.75S, 179.75E (centre). We have no longer provide a plain text format file as well. Please request this only if you cannot progress without it. The data run from January 1903 to December 2025, driven by the CRU-TS v4.10 monthly climate dataset which is available from here: https://crudata.uea.ac.uk/cru/data/hrg/ IMPORTANT CHANGES IN THIS VERSION This version (data ending in 2025) of scPDSI has been created with a new python code base generated by porting the older Fortran code to python. All prior versions (data ending in 2024 or earlier) were created using the old Fortran scPDSI code. The initialisation/spinup procedure is also different in new process, causing the first two years (24 months) to be unduly affected by spinup artefacts (a very small number of implausibly large scPDSI values). Therefore the scPDSI files have been truncated to begin in 1903. We intend to improve the spiniup procedure in future versions to allow us to provide files beginning in 1901 once more. Changes to the Water Holding Capacity dataset have affected which grid cells have data: 372 grid cells were missing in previous version but now have data in v4.10, while 6551 grid cells (e.g. some islands, especially Arctic islands) had valid values in previous version but are now set to missing in v4.10. CAVEATS and NOTES: 1. Values during 1901 are unreliable because scPDSI was initialised in 1901 and no CRU-TS climate data were available for 1900. Currently, values in 1901 and 1902 are not provided, to avoid unreliable values being used. 2. In CRU-TS, grid cell climate values are relaxed towards the 1961-1990 mean climatology when there are no or few weather station observations within the Correlation Decay Distance (CDD). The CDD is 1200 km for monthly temperature but only 450 km for monthly precipitation. This means that the scPDSI of grid cells may be closer to zero (i.e. normal conditions), underestimating the occurrence of anomalously dry or wet spells, in regions and times with poor data coverage. 3. The underestimation of anomalously dry or wet spells (see note 2) will affect the early decades of the twentieth century the most, but it can also occur during recent years because not all observations are easily available in near-real time for updating CRU-TS. A rough indication of the scale of this limitation is given by the number of precipitation observations incorporated into the preliminary version of CRU-TS v3.25 (the version used for the 1901-2016 scPDSI dataset). For the 1990s and 2000s, approximately 5,000 observations were available per month. For 2011-2015, the mean number of precipitation obserations per month in CRU-TS v3.25 was 2200, 2166, 2064, 1725 and 1676, respectively. 4. The scPDSI data values in 2025 are based on CRU-TS v4.10 created in March 2026, and future versions will likely incorporate additional observations for 2025 that were not available to us so early in 2026. When using these data, please cite: Barichivich J, Osborn TJ, Friedman AR, Harris I, van der Schrier G and Jones PD (2026) Monitoring global drought using the self-calibrating Palmer Drought Severity Index [in "State of the Climate in 2025"]. Bull. Amer. Meteor. Soc., 107, S71-S72 https://doi.org/10.1175/BAMS-D-26-0141.1 and van der Schrier G, Barichivich J, Briffa KR and Jones PD (2013) A scPDSI-based global data set of dry and wet spells for 1901-2009. J. Geophys. Res. Atmos. 118, 4025-4048 (10.1002/jgrd.50355). The publication which introduced the self-calibrating PDSI, and gives a large amount of detailed information about its calculation, is: Wells, N., Goddard, S. and Hayes, M. J. (2004) A Self-Calibrating Palmer Drought Severity Index, J. Climate 17, 2335-2351 Support for the 2025 update was provided by Horizon Europe research and innovation programme (ERC-starting grant CATES, grant agreement No. 101043214) and the NERC National Centre for Atmospheric Science (NCAS).