Detection and inventory of saline seep using color infrared aerial photographs and video image analysis

Saline seep is an important conservation problem affecting dryland agriculture in the northern Great plains of North America. Surveys on the location and acreage of saline seep are needed to assess economic losses and to develop strategies for reclamation. Survey needs have been met with varying success by such remote sensing techniques as photography, thermo-graphy, and satellite imaging. However, high costs, sensor resolution limits, and lack of ground reference data prevent a fully operational remote sensing system to survey saline seep. Small-format photographs (1:32,000 scale) were taken to record saline seeps in a 200 square mile portion of the Sage creek watershed in Liberty County, Montana. Beginning, developing and mature seeps were visually identified and subsequently mapped and inventoried using a video image analyzer. Ground data on soils, crops and weeds were used to verify the interpretations. Beginning, developing and mature seeps were detected on the 1:32,000 scale color infrared photographs. Optimal image date for recording seeps occurred when rank growth, darker color, and delayed crop maturity of beginning seeps appeared in fields that were otherwise nearly ripe and ready for harvest. Developing and mature seeps were visible on CIR imagery for most of the growing season. Electromagnetic and electrical conductivity measurements of salt content and depth distributions confirmed that extra water caused the heavy crop growth on beginning seeps. Developing seeps were areas where crops were reduced or eliminated due to high salinity. Mature seeps were areas with salt crusts, halophytic weeds and very high salinity. The inventory revealed 2.4 percent of total cropland was seep-affected. However, many individual farms were as much as 7 percent affected. Mapping accuracy was 85 percent for mature seeps and 76 percent for beginning seeps: Beginning, developing and mature seeps are mappable from small-scale CIR imagery. Crop expression of salinity is the key to detecting saline seeps. The 2 week period ending about 1 week before harvest is the best time for recording beginning seeps. Video image analysis is accurate and rapid for quantifying, compiling and sorting large amounts of map data. An operational survey is possible pending training programs and adaptation of interpretation guidelines to fit other areas. DETECTION AND INVENTORY OF SALINE SEEP USING COLOR INFRARED AERIAL PHOTOGRAPHS AND VIDEO IMAGE ANALYSIS by Daniel Sinpson Long A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in