Kenya: Ecosystem Based Adaptation for Resilience to Droughts in Kenya’s Arid and Semi-Arid Rangelands

Here, we summarise progress toward an improved methodology for economic and financial assessment, compared to the original assessment in 2019. The triple dividends of resilience-building (Tanner et al., 2016; Wilkinson & King-Okumu, 2019; Heubaum et al., 2022; Yokomatsu et al., 2023; van Zanten et al., 2023; Cuevas et al., 2024), include: 1) avoiding losses; 2) securing economic benefits thanks to reduced risks; and 3) social, environmental and economic co-benefits associated with the investments. We draw on monitoring systems now put in place through the project, in conjunction with supplementary mapping, modelling and valuation techniques to estimate the project contributions to ecosystem service values and economic growth. Rangeland handbooks (Herlocker et al., 1993; Schwartz et al., 1991; Shaabani et al., 1992), and other sources helped to characterise relationships between climate, water and land management, vegetation, livestock productivity and value chains.

Watershed Organisation Trust (WOTR) developed analysis of changes in landcover visible over the period 2011-2024 in Landsat 5, Landsat 8, and Sentinel-2 satellite images from March to June for each year (Figures 1-3). They adapted a regression model (from Onyango, 2015) to estimate the volume of biomass production in 3 wards in the study area based on the Normalized Difference Vegetation Index (NDVI) and standard deviation (Figure 2). This showed the overall increasing trend in biomass production, interrupted by decreases in biomass production during the drought periods in 2011 and 2022. It is notable that although the more recent drought began in 2021, the effects on the vegetation conditions only became pronounced in 2022. 

According to past weather patterns, extreme drought events occur in Kenya approximately once in a decade (with previous notable droughts spanning 2008-11 and then 2018-2021). Intercommunal tensions spiked at that time and the circulation of small arms escalated. A summary comparison of the effects of droughts on Kenya’s Gross Domestic Product (GDP) seems to show increased resilience over the decade preceding the project. A detailed review of statistics from the counties was cancelled due to withdrawal of USAID support for assessing the costs of the 2022 drought. We reviewed scope to quantify and assess the benefit streams from improving water and rangeland management interventions in terms of volumes of rainfall and runoff conserved in the restored areas, as captured in water pans, recharge to the subsurface (accessible by shallow wells) and aquifers (accessible by deeper wells in some areas) and rooftop water harvesting (WRA, nd). Recharge areas were identified across the water catchment.

On the basis of past climate patterns, we could conservatively estimate that the main pay-back for the GCF investment in drought resilience within this project would be during the next severe drought episode which might occur in 2028-31, based on current patterns. Therefore, our assessment must consider two timeframes: returns on investment in terms of benefit flows to households and stakeholders during the project timeframe (2021-27), and macro-economic returns over longer time horizons up to 2031 and 2050. We prepared with- and without-project scenarios for effects on access to water, vegetation, livestock, the economy and public expenditures (Table 1). We then identified methods to model and value these effects over past and future years (Table 2). 

To understand the returns that could be anticipated within the project lifetime, we predict a linear increase in ecosystem service benefit flows 2024-7 (Table 3). We then explore a longer-term scenario in which a major drought onset would emerge 2028-31 (if not sooner). As yet, we did not include use of available climate models and scenarios which may suggest an earlier onset, and nor do we assess the effects of significant cuts to previous drought resilience investments through other sources. 

Project information confirmed significant elasticity in the economic value of key ecosystem services affected by EbA (water, food, gums and resins), depending on levels of investment in value chains and local service provision (CI, 2024; COMSTRAT, 2024). This reconfirms previous observations of significant untapped value for private sector and the national economy in the pastoral production systems (King-Okumu, 2015; Lutta et al., 2020; Thomas et al., 2024). We also know that value generated through regulating services, such as groundwater recharge, storage and availability during droughts, continues to multiply land values over time (King-Okumu, 2018). These work together with economic productivity and growth across all sectors to grow the economy, generate public revenue and reduce future drought risks.

Without inclusion of the full value of rangeland management effects on water at the catchment level, an initial assessment suggested that the project could not demonstrate a positive return on investment unless payments for carbon sequestration were part of benefit calculation. However, ineffective governance and an illegal use of firearms made the system for carbon payments inaccessible to affected communities (Mukpo, 2025). We therefore revised the initial assessment to increase consideration of the other ecosystem service values and value chains. The more complete evaluation of changes in ecosystem service benefits could already demonstrate a positive rate of return within the project lifetime and beyond -with or without the payments for carbon.

Using information available on the number and volumetric capacity of the water cisterns water pans and wells (GOK, 2023; GoK, 2023b, 2023a; Jarso et al., 2017) enables us to identify better returns in light of increased volumes of water made available each year thanks to improved siting, maintenance and management of waterpoints and land in the rangelands (King-Okumu, 2016; 2016). Excel supports prediction of simple linear increases in the volumes and values of ecosystem service flows, including water availability, gums and resins, fodder, livestock, meat and milk production following improvements supported by the project in the project area (Table 3). These could be compared to the actual and potential value of carbon sequestration under different institutional and market conditions. 

In absence of Geographic Information Systems (GIS) connecting the rangeland resource monitoring and management plans into the national drought management system, assumptions borrowed from two previous studies (Gies et al., 2014; Muriuki, 2023) proved helpful to explore the effects of rangeland management on groundwater recharge across the wider catchment area. These methods also showed potential to accommodate varying future climate scenarios and to feed relevant values into an economic assessment. This would merit discussion with experts at Kenyatta University (KU), Texas A&M University and the United States Department of Agriculture Agricultural Research Service (USDA ARS).

Reflecting on available understanding of the macro-economic effects of disasters (GFDRR, 2014; Yokomatsu et al., 2023), we identified further needs to supplement available guidance on appraisal of Nature-based Solutions (NbS) (King-Okumu, 2017; Van Zanten et al., 2023; Thomas et al., 2024) in light of the cascading risks, impacts and economic opportunities associated with the economics of resilience to droughts in Kenya (GoK, 2012; Cabot-Venton, 2018; Cabot Venton, 2019; Venton et al., 2019). A Computable General Equilibrium (CGE) modelling approach is favoured for some macro-economic assessments (Bazzana et al., 2022). These models depend on input values collected in a Social Accounting Matrix (SAM) (Managi et al., 2024). GIS gaps remaining could be filled through ongoing coordination by NDMA and partners in TWENDE with work already underway to improve national natural capital accounting systems (KNBS, 2024).