Using SWAT to predict crop yields and salinity levels due to future weather scenarios in the North Fork watershed in southwest Oklahoma

Estimating future flow and salinity levels within stream systems and crop yields based on future climate conditions is an increasingly important issue. Saline irrigation water is a major concern in the North Fork of the Red River watershed in southwestern Oklahoma. The Elm Fork Creek flows through salt deposits, making the creek and its receiving stream, the North Fork River, too saline to use for irrigation. This greatly reduces the number of hectares that can be utilized for agricultural crops within the watershed. Future weather scenarios were analyzed to determine how varied precipitation and temperature would impact salinity levels in the streams and crop yields within the watershed. Due to no significant weather trends in the 57 year record within the watershed, WXGEN was used to simulate a potential 10% increase and decrease in precipitation from 2011-2060. Output from the weather generator was input into a calibrated SWAT model to simulate wheat and dryland and irrigated cotton yields for the various weather scenarios. Using an existing correlation between ionic strength and streamflow, salinity levels were estimated and found to increase by 9% for the dry scenario and decrease 31% for the wet scenario at the USGS gage station located downstream from salt springs on the Elm Fork River. The levels ranged from 22 dS m^-1 to 35 dS m^-1, which were much higher than the levels (10 dS m^-1) where only very tolerant crops can be successfully grown. Though there was not a significant difference in wheat yields, the simulations with a 10% increase in precipitation had significantly higher yields than the simulations with a 10% decrease in precipitation for both the dryland and irrigated cotton.