Predicting Minnesota Rainfall Using Atlantic Ocean Salinities

Under the high pressure cells of the subtropical oceans, evaporation exceeds precipitation by 1-3 meters/year. Thus, these areas are where the ocean exports significant amounts of latent heat energy and water to other regions of the climate system (at a rate of 75W/m² per m/year of net water loss). Latent heat fluxes from the ocean are an order of magnitude larger than sensible heat fluxes and significantly larger than long wave radiation, and thus represent the atmosphere's main supply of energy from the ocean. The oceanic signature of the large water losses is seen in elevated salinities; the largest open ocean salinities are always found at the surface in the centers of the gyres under the subtropical high pressure systems. Well documented long-term trends in ocean salinity, with salty areas getting saltier and fresh areas getting fresher, show that salinity is a sensitive indicator of change in the water cycle. This is to be expected since the global water cycle is thoroughly dominated by its oceanic component; e. g., oceanic evaporation exceeds the flow of all rivers by more than an order of magnitude. This strong signal motivated the search for teleconnections from sea surface salinity (SSS) in the subtropical gyres to rainfall on land. Here it is shown that seasonal surface salinity variability is indeed useful as a predictor of terrestrial rainfall. Surprisingly, it is the northern portions of the subtropical gyres that provide the strongest teleconnections. The winter and spring SSS in the northeast sector of the North Atlantic subtropical gyre is a good predictor of summer rainfall in the Sahel region of Africa. The winter and spring SSS in the northwest sector of the gyre proves to be a good predictor of summer rainfall in the upper Midwest and Great Plains. Analysis shows that soil moisture is the likely mechanism for the one-season delay. It interacts with the low-level jet to bring moisture northward in the central US. The SSS indicators described here are independent of El Nino and the North Atlantic Oscillation and prove to be better predictors of terrestrial rainfall than sea surface temperature. SSS is now being monitored by the Aquarius satellite and soil moisture by SMAP, both L-band radiometers. With the ARGO float program providing quality in–situ salinity data, significant improvements in seasonal rainfall forecasts appear to be within reach.