Measuring changes in areal extent of historic wetlands at Great Sand Dunes National Monument, Colorado 1936-1995

Abstract

Great Sand Dunes National Monument (GRSA) is a unit of the National Park system in south central Colorado. With funding from the Colorado Historical Society, a series of studies were undertaken for an environmental history of the GRSA area and the San Luis Valley. Park managers were concerned over the disappearance of small wetlands in northwestern GRSA. The objective of this study is to document and analyze the changes to the wetlands through the study of digital, georeferenced images and to relate these changes to climatic and hydrologic factors. Ten sets of aerial photographs were obtained, from 1936 to 1995, with at least one set from each decade except for the 1940s. All photos were scanned into a digital format. A system was devised to mosaic the images prior to rectification due to the lack of ground control in the area. Land cover was digitized from the mosaics including the wetlands and sand type. Size and spatial distribution of the wetlands were analyzed. Analysis shows that the greatest total number of wetlands and acreage totals were present in the 1936 and 1937 photo sets. In 1937, 114 wetlands were found, 47% have water at the surface. By 1953 the total number of wetlands dropped to 38 and by 1975 only 22 remained, with only 1 having water at the surface. The total number of wetlands has increased in recent years primarily due to sub-irrigated meadows. A large increase in the vegetation cover has .occurred since 1936 to the present, increasing from 20% of the area in 1936 to 47% by 1995. Climatic data were collected to analyze possible causes of the changes to the wetlands but the study was limited by the lack of long-term data. Weather data is consistently available since 1948. Well data is of short term and sporadic nature. Two sources of long-term data were available. The discharge of the Rio Grande has been monitored in Del Norte, CO since 1906. A dendrochronology study was done in the area of GRSA in 1980. The precipitation data, discharge and dendrochronology data were summed, averaged or offset for monthly or annual intervals prior to the dates of the aerial photo sets. These values were correlated with the acreage of the wetlands for each of the photo years by means of linear regression. Very poor correlation resulted between the precipitation indices and wetland acreage. A surprising result came from correlation of the wetland acreage with the dendrochronology and discharge data. Over 58% of the wetlands variation can be explained by the 15 to 19 year offsets of dendrochronology data, 55% can be explained by the 20 year offset of Rio Grande discharge. These results are interpreted that the offset indices are related to the slow change in total area of the wetlands rather than fluctuations in the water table.