Targeted climate-smart irrigation advisory using integrated climate indicators

Increasing rainfall variability and rising temperatures are intensifying irrigation demand across western Kenya, particularly in smallholder systems in the Lake Victoria Basin. Yet many existing tools rely on single-variable indicators, limiting their usefulness for integrated and anticipatory decision support. This report develops a Compound Irrigation Advisory Index (CIAI) to support Targeted Climate-Smart Irrigation Advisory (T-CSIA) in Kisumu, Siaya and Homa Bay counties by combining a Precipitation-based Irrigation Advisory Index (PIAI) and a Temperature-based Irrigation Advisory Index (TIAI) derived from high-resolution daily CHIRPS precipitation (1981–2025; ~5 km) and CHIRTS temperature data. The PIAI integrates precipitation concentration, a length-of-growing-period proxy based on rainy days ≥5 mm, interannual variability (coefficient of variation) and long-term rainfall trends (Theil–Sen slope), while the TIAI captures temperature-driven irrigation stress using reference evapotranspiration (Hargreaves–Samani), climatic water deficit, evaporative imbalance (ET₀–P), aridity (PET/P), growing degree days and heat-stress days. Component indicators are normalised to a 0–1 risk scale and combined under a water-balance rationale, with seasonal probabilistic rainfall and temperature forecasts from ICPAC incorporated to generate forward-looking advisories with approximately one-month lead time. Results show strong spatial heterogeneity in irrigation need, with eastern Kisumu consistently emerging as a hotspot under both historical and forecast-integrated conditions, while Siaya generally exhibits lower relative stress and Homa Bay displays mixed moderate-to-high patterns. Forecast integration primarily intensifies historically vulnerable zones rather than shifting the geographic distribution of risk and high CIAI classes align with locations where farmers report practising irrigation; comparison with the Advanced Drought Response Index (ADRI) further indicates consistency with areas experiencing mild to moderate agricultural drought.