Abstract ID: 3.9211 | Accepted as Poster | Poster | TBA | TBA

Mount Kilimanjaro, the highest peak in Africa (5895 m), rises from the lowland savannah of northern Tanzania. Being near the equator, it includes a wide range of ecological zones from tropical savannah to ice/snow (nival zone) near the summit. Kilimanjaro therefore serves as a critical indicator of climate change in equatorial regions. Studies have found increased warming on Kilimanjaro at higher elevations, known as elevation-dependent warming (EDW). These temperature changes lead to melting snow and shrinking glaciers, threatening the unique ecosystems and water resources. However, due to the sparse distribution of high-elevation weather stations, there is limited understanding of temperature patterns and the drivers of change. One approach to address this limitation is the use of satellite-derived land surface temperature (LST), which has a strong relationship with air temperature (Tair) and offers good coverage even in complex mountainous terrain. However, LST and Tair differ in their physical definitions, measurement methods, and responses to environmental conditions. In complex mountainous regions like Kilimanjaro, LST alone cannot well model Tair. Understanding how environmental factors influence the LST-Tair relationship is therefore essential for improving temperature models and assessing EDW. This study examines the modelling of Tair from Moderate Resolution Imaging Spectroradiometer (MODIS) LST on Mount Kilimanjaro from 2004 to 2024, incorporating elevation, topography, land cover (particularly vegetation and snow) and solar radiation as auxiliary predictors. In situ Tair data from 26 sites, covering elevations from 991 m to 5803 m is used to develop and validate models, and discuss the environmental factors influencing the LST-Tair relationship. Our results show that the LST-Tair relationship is more variable during the day, with a pronounced positive temperature difference (ΔT, LST minus Tair). Additionally, ΔT increases at higher elevations. Finally, temperature trends based on the calibrated LST data are analysed to determine whether EDW is evident and how trends vary with elevation, aspect, vegetation and snow.