Can Hyperscale Data Centers Alter Local Meteorology in the Kathmandu Valley? A Simulation-Based Assessment

Sujan Bhattarai , Saroj Prasad Mainali, Suman Bhattarai

There is growing interest in Nepal in hosting large data centers inside the Kathmandu Valley, but how the valley’s atmosphere would respond to substantial additional heat emissions remains uncertain. The Kathmandu Valley is a closed bowl-shaped basin surrounded by mountains, with persistent nocturnal cold pools and weak boundary-layer winds that limit the dispersal of pollutants and heat from the basin. These features make it a difficult receiving environment for a large data center, whose operation produces huge quantities of waste heat that must be released into the local atmosphere. This study presents a scientific assessment of whether siting such a facility in the valley would cause measurable temperature changes in the city, intended to scientifically inform the ongoing public discussion. We use publicly available satellite and reanalysis datasets, including ERA5-Land, MODIS, Sentinel-2, ESA WorldCover, GHSL built-up, and SRTM elevation data, to estimate the additional heat that a hypothetical 50, 150, or 500 megawatt data center would inject into the valley atmosphere if placed on the southern valley floor. The analytical results show that the temperature impact during daytime would be small, because afternoon up-valley winds advect the heat plume away from the urban core. At night, however, drainage flow reverses the wind direction and carries the plume toward central Kathmandu, raising local temperatures by approximately 0.1 to 0.5 °C for a medium-sized facility and 0.4 to 2.0 °C for a very large one. Based on these results, a small data center on the order of 50 megawatts is of limited meteorological concern under the valley's regime, while larger facilities produce a temperature signature that becomes measurable at the urban core. While this study evaluates atmospheric implications, the broader environmental feasibility of hyperscale data-center deployment depends heavily on freshwater use and wastewater management. It remains unclear how such demand, along with cooling-related wastewater discharge, would be managed. These hydrological implications therefore require dedicated assessment before any large-scale data-center development is approved.