Mapping the severity of soil salinity in the Red River Valley
A major concern in many regions is that the sustainability of food production is threatened by the build-up of salts in agricultural soils. One source of concern is the many examples of historical societies that have succumbed to soil salinity. But information on the extent and severity of soil salinization is often unreliable and subjective. More robust and rapidly deployable methods to quantify soil salinity would greatly help to identify and manage local problems, as well as to assess the global threat from soil degradation.
To this end, we work on developing and testing methods to remotely sense soil salinity. One project focuses on the Red River Valley in the central United States, an area that is thought to be among the most saline in the country. In an effort to develop techniques that can rapidly and inexpensively assess salinity hazards, we rely on using MODIS satellite data as illustrated in the figure above. Our approach attempts to correlate vegetation condition averaged over many years with ground-based measurements of salinity. We have found that by averaging over 5+ years, the ability to identify high saline areas improves relative to single year measures. This is because many factors which obscure the signal in a single year, such as weather or management differences across space, tend to dampen out over the long term.
A related project is the Mexicali Valley in Baja California, Mexico. Again, this region is thought to have major salinity problems, but monitoring of broad scale patterns has proven difficult. Here we have used higher resolution Landsat imagery to map wheat yields, and then focus on consistent patterns across the multiple years. In this case, we have processed imagery from each growing season from 2000-2009 for a total of 10 years.