H31F-1496
Watershed-scale Evapotranspiration Changed Little over 50 years of Agricultural Land Abandonment in Southern Michigan
Wednesday, 16 December 2015
Poster Hall (Moscone South)
Stephen K Hamilton, Michigan State Univ, Hickory Corners, MI, United States, Mir Zaman Hussain, Michigan State University, Great Lakes Bioenergy Research Center, East Lansing, MI, United States and Christopher J Lowrie, Michigan State University, Computational Mathematics, East Lansing, MI, United States
Abstract:
The difference between precipitation and stream discharge over annual periods provides an indication of the total water loss to evaporation and evapotranspiration. The response of evaporative water loss to land cover change affects groundwater recharge, stream flow, and lake levels. This study examined the watershed water balance for Augusta Creek, which drains a 95-km2 glacial landscape in southwestern Michigan covered by cropland, grassland, forest, and wetlands. The climate is humid and temperate; between 1964-2014 the water-year precipitation averaged 948 mm and ranged from 695-1386 mm with no temporal trend. Over the study period the percentage of land in agriculture has decreased to about a third of its original extent, with abandoned lands gradually transitioning from old fields to woody vegetation. Comparison of precipitation on the upland watershed to baseflow discharge (USGS data; baseflow estimation by WHAT model) across the 50-year record shows that total evaporative water loss averaged 563 + 103 mm and ranged from 385-897 mm, with no apparent trend over the record. The evaporative water loss accounts for a mean + s.d. of 59 + 6% of precipitation (range, 48-70%). Evaporative water loss was positively related to total precipitation (r2 = 0.74. These results are interpreted using a Budyko plot framework to facilitate comparison with other settings. This water balance approach to infer evaporative water loss compares well with direct measurements in the same watershed since 2009 using eddy covariance (grasslands and crops) and soil moisture monitoring by time-domain reflectometry (grasslands, crops, and forest). Thus the evaporative water loss, which is predominantly by evapotranspiration, has been remarkably similar across a period of changing land cover, leaving a relatively consistent proportion for groundwater recharge and streamflow.