From Past to Future: Uzbekistan’s Climate Signals Through Time

Abstract. This study investigates long-term trends and future projections of key climate variables, evapotranspiration, land surface temperature (LST), normalized difference vegetation index (NDVI), soil moisture and precipitation, across Uzbekistan using multi-source satellite and reanalysis datasets within the Google Earth Engine (GEE) platform. Spanning the period from 1995 to 2024, the analysis applied the Mann-Kendall test to assess the statistical significance of observed trends, revealing significant increases in LST, NDVI and evapotranspiration, while trends in soil moisture and precipitation were statistically insignificant. To forecast future trajectories (2025-2050), the autoregressive integrated moving average (ARIMA) model was employed, indicating continued warming, vegetation growth and rising evapotranspiration, with marginal changes in precipitation and a possible decline in soil moisture. Model performance was evaluated through a 70/30 training-test split, where NDVI achieved the highest R2 (0.64), followed by precipitation, LST, evapotranspiration and soil moisture. These results suggest ARIMA can capture temporal patterns to a degree, but more extensive datasets and integrated models may be necessary for higher accuracy. To sum up, findings point to a warming and drying climate scenario, underscoring the urgency of proactive land and water management strategies to ensure ecological and agricultural sustainability in Uzbekistan under evolving climatic conditions.