During Northern Hemisphere summer, strong surface winds parallel to the eastern coast of Africa (the Somali jet), cause widespread upwelling and cooling in the Indian Ocean. Coastal waters are up to 5° C colder than open ocean at the same latitude, and upwelling sea surface temperature effects can extend as far out as 12 degrees (~1350 km) from the coast. SST data sets used in global, and even regional, models do not capture this structure in the SST field. But recent satellite observations from the AVHRR MCSST project provide high-resolution SST data sets that can be used to assess the influence of coast upwelling on the Indian monsoon.

Mesoscale model (MM5) simulations are conducted using high-resolution (74 km) and coarse resolution (274 km) SST data sets to understand how the cool SSTs associated with coastal upwelling influence the Indian monsoon. MM5 was modified and tested for a domain centered on the North Indian Ocean, and produces a simulation of the Somali jet, East African precipitation, and the Indian monsoon that is more realistic than those generally found in global model simulations.

When higher resolution SSTs are used, the monsoon simulation over India changes. The Somali jet shifts northward over the Arabian Gulf due to the enhanced zonal temperature gradient and this leads to changes in the locations of moisture convergence along northwestern Indian coast, as well as increased coastal precipitation late in the season.