2.3.1. Rainfall (P) Only Products

The data from the Global Precipitation Climatology Centre (GPCC), operated by the German Weather Service, consists of the world's largest database of station-based precipitation data [44]. The primarily monthly data is used to develop gridded products such as version 6 monthly rainfall data, which integrates the largest station number. The GPCC data showed reliable performance when compared at various locations at the global level compared to the Climatic Research Unit gridded data (CRU CL 2.0) and ERA40 product from the European Center for Medium-Range Weather Forecasts (ECMWF). The data from the Global Precipitation Climatology Project (GPCP) from the World Data Center for Meteorology, in turn, is a monthly gridded product built by merging satellite estimates and gauge analysis from the GPCC. Version 2.3 includes adjustments for improved rainfall estimates compared to version 2.2 [45]. A study over the complex terrain of the Ethiopian highlands by Dinku et al. [7] showed the applicability of the product under those circumstances compared to the TRMM 3B43 and the Climate Prediction Center (CPC) Merged Analysis of Precipitation (CMAP) data.

The CHIRPS data is a merged product including five satellite-based and groundstation products [41]. It has previously proven reliable in the uneven topography of East Africa [30]. Over Kenya, the product has demonstrated remarkable performance as Ayugi et al. [46] found out and over drier regions as Gebrechorkos et al. [35] report, where it outperformed the Africa Rainfall Climatology (ARC2) and CHIRPS datasets. The latest version of TAMSAT data (TAMSAT 3.1), in turn, merges Meteosat thermal infrared imagery and rain gauge observations covering the entire African continent since 1983 [47]. Alongside the TRMM 3B42 and Climate Prediction Center Morphing Method (CMORPH) product, TAMSAT demonstrated high performance over the complex Ethiopian highlands in a study by [7]. Another largely satellite based product, the PERSIANN-CDR (Climate Data Record), is developed from GPCP and satellite-based data [48]. The PERSIANN-CDR has proven useful in detecting disasters as Ashouri et al. [48] showed in the 2005 Katrina hurricane product verification study, comparing also GPCP, TRMM, and the CPC gridded data.

#### 2.3.2. Rainfall (P) and Temperature (Tmin/Tmax) Products

The Kenya Meteorological Department (KMD) indicated Machakos and Makindu, located approximately 100–200 km away from the study area KMD [32], as the two nearest ground stations. The nearest station, Kitui Agrometeorological Station, had only a 5 year record and too many data gaps to be useful for our analysis. The same applied to adjacent volunteer stations [32]. Hence the gridded data products could only be compared to the Machakos and Makindu stations that had reliable records [32,49]. The gridded products are summarized in Table 1. The KMD also provided gridded data for Kitui West [32,50]. This product is developed through the Enhancing National Climate Services (ENACTS) program [50–52], which works with national meteorological services across Africa to improve the quality of climate data and enhance access in essential sectors such as agriculture to counter the problem of scarce ground-based stations [29,52]. The KMD product combines spatially downscaled reanalysis data and bias corrected satellite-based rainfall estimates with sparse station-based observations. For Tmax and Tmin, 37 weather stations across Kenya were used and merged with data from the JRA-55 (Japanese 55-year Reanalysis) product (see Table 1 for JRA-55 background) [53]. Rainfall was generated using data from about 700 stations which were merged with satellite data from the CHIRPS product (see Table 1) [41,50].

The CRU TS data is a gridded product based on angular distance weighting of groundstation data from national meteorological services around the world [54]. The product's performance has been compared to the GPCC data. The JRA-55 data, produced by the Japanese Meteorological Agency, is an improvement of the predecessor, JRA-25, where shortcomings, such as cold bias in the lower atmosphere, dry bias in the Amazon, and a longer time scale, since 1958, have been addressed [55]. Following Hua et al. [56], the product has demonstrated reliability in central equatorial Africa where a comparison was made with other reanalysis products including MERRA-2, ERA-Interim, The Twentieth Century Reanalysis (20CR), the Climate Forecast System (CFSR), the National Center for Atmospheric Prediction NCEP-1 and NCEP-2. The ERA5 data is a fifth-generation reanalysis product of the ECMWF [57]. It has a longer temporal coverage and higher resolution than the predecessor, ERA-Interim, and provides more parameters at hourly resolution accompanied by uncertainty information. A study by Tetzner et al. [58] compared the performance of the product to in-situ stations, with Kawohl [59] revealing the usefulness of ERA5 especially at high elevations. The MERRA-2 data is a reanalysis product of the Global Modeling and Assimilation Office of the Goddard Space Flight Center developed

towards the aim of an integrated earth system analysis [60]. The satisfactory performance of the product as compared to the Global Precipitation Climatology Project (GPCP) and JRA-55 products is depicted by Bosilovich et al. [61] and by Hua et al. [56] over central equatorial Africa through comparison with the new gauge-based NIC31 product alongside other reanalysis data such as JRA-55 and ERA-Interim.

**Table 1.** Rainfall (P) and temperature (Tmin/Tmax) products used in the computation of SPEI. Original daily data were aggregated to a common monthly resolution. Only validated and widely used products with a length of more than 30 years were used.



**Table 1.** *Cont.*

#### *2.4. Areal Averages*

The meteorological data were averaged over the study area by weighted average, proportional to the contribution of each grid cell to the study area shape (see Figure S1 and Equation (S1) of the Supplementary Information). For each data product, the grids differed in their intersection with the study area (see Figure S2). Correlations of the areal averages with the nearest ground-stations at Machakos and Makindu and the gridded rainfall data provided by the KMD were greater than 0.6 (see Figure S3). Following Sun et al. [63], we used the native resolution of the products (Table 1) in the computation of areal averages. Nevertheless, topographic information could be included in interpolation in future studies, certainly when covering greater areas in East Africa where the topography is highly variable.

#### *2.5. Key Informant Interviews*

Balint et al. [3] recommend the triangulation of SPEI output in order to reinforce the results while also contributing to a broader understanding of the temporal evolution of droughts and ongoing responses. Following Denscombe [64], we additionally view methodological triangulation (referred to as triangulation in the text) as an optimal approach for integrating qualitative and quantitative data to generate a confirmatory picture. Therefore, in addition to the SPEI calculations using the 40 blends of rainfall and temperature products, information on drought occurrence and severity was obtained by interviews with 14 key informants with a track record of working on droughts and related activities, e.g., food security, humanitarian, and farm-based interventions, in the study region (Table 2). The key informants include representatives from Non-Governmental Organizations (NGOs) and government agencies at the national and county-level). The interviews were conducted between August 2020 and February 2021 as video meetings and were preceded by official communication. They included discussions under the broad subjects of drought frequency, trends, and history as observed in the interviewee's line of activity, nature of responses implemented with regard to water storage and on-farm interventions, collaboration with the affected communities, and experiences and prospects under the relatively new county governance system. The interview guide is included in the Supplementary Information under Breakdown S1. A snowball sampling approach was used, where each key informant was asked to suggest equally active organizations in the study area for further interviews [64]. The organization's profile and activities were also reviewed via desktop-based research. Some interviews were recorded upon consent of the interviewee; for others, notes were taken.


**Table 2.** Key informants with operations in Kitui West and their corresponding designation, categorization, organization, or department and interview date.
