Comparing CMIP-3 and CMIP-5 climate projections on flooding estimation of Devils Lake of North Dakota, USA

BackgroundWater level fluctuations in endorheic lakes are highly susceptible to even slight changes in climate and land use. Devils Lake (DL) in North Dakota, USA is an endorheic system that has undergone multi-decade flooding driven by changes in regional climate. Flooding mitigation strategies have centered on the release of lake water to a nearby river system through artificial outlets, resulting in legal challenges and environmental concerns related to water quality, downstream flooding, species migration, stakeholder opposition, and transboundary water conflicts between the US and Canada. Despite these drawbacks, running outlets would result in low overspill risks in the next 30 years.MethodsIn this study we evaluated the efficacy of this outlet-based mitigation strategy under scenarios based on the latest IPCC future climate projections. We used the Coupled Model Intercomparison Project CMIP-5 weather patterns from 17 general circulation models (GCMs) obtained under four representative concentration pathways (RCP) scenarios and downscaled to the DL region. Then, we simulated the changes in lake water levels using the soil and water assessment tool based hydrological model of the watershed. We estimated the probability of future flood risks under those scenarios and compared those with previously estimated overspill risks under the CMIP-3 climate.ResultsThe CMIP-5 ensemble projected a mean annual temperature of 5.78 °C and mean daily precipitation of 1.42 mm/day; both are higher than the existing CMIP-3 future estimates of 4.98 °C and 1.40 mm/day, respectively. The increased precipitation and higher temperature resulted in a significant increase of DL’s overspill risks: 24.4–47.1% without release from outlets and 3.5–14.4% even if the outlets are operated at their combined full 17 m3/s capacity.DiscussionThe modeled increases in overspill risks indicate a greater frequency of water releases through the artificial outlets. Future risk mitigation management should include providing a flood warning signal to local resource managers, and tasking policy makers to identify additional solution measures such as land use management in the upper watershed to mitigate DL’s flooding.

Evapotranspiration Mapping with METRIC to Evaluate Effectiveness of Irrigation in Flood Mitigation for the Devils Lake Basin

A period of excess precipitation since 1993 in the Devils Lake basin in northeastern North Dakota has caused extensive flooding of agricultural land and has raised the question of whether irrigation of agricultural crops to increase evapotranspiration (ET) might be an effective way to remove water from the basin. The objectives of this study were to compare ET estimates derived from application of the Mapping ET at High Resolution with Internalized Calibration (METRIC) algorithm for North Dakota conditions (METRICND) under irrigated and rainfed conditions and to assess the potential for irrigation to increase crop ET as a flood mitigation strategy. Weather data, land use maps, and Landsat 5 Thematic Mapper imagery from 2006, 2007, and 2008 were used as inputs to the METRICND model. The ET for irrigated crops (ETIrrigated) was estimated at five test sites from the Devils Lake Basin Water Utilization Test Project (DLBWUTP). The ET for the predominantly rainfed study area (ETRainfed) was estimated using land use maps to identify locations of the same crops as were present on the test sites. The METRICND model was compared to ET values derived from an eddy covariance (EC) system for approximately two months in 2007 at an irrigated alfalfa test site in the DLBWUTP; the mean absolute error between METRICND and the EC system for the comparison period was 0.51 mm d-1. Linear regression of ET (in mm) for the test sites and the larger study area yielded ETIrrigated = 1.23 × ETRainfed + 4.77 with R2 = 0.96, and a t-statistic indicated that the slope was greater than 0 at p = 0.001, indicating the potential for increased ET under irrigation. However, addition of large volumes of irrigation water to the predominantly poorly drained soils in the basin will cause waterlogging and trafficability problems. Installation of subsurface drainage may help alleviate waterlogging, improve crop productivity, and increase ET, but subsurface drainage brings its own complications of disposal of the drained water, salinity of the drainage effluent, and possible sodicity problems on some soils. Keywords: Drainage, Evapotranspiration mapping, Irrigation, METRIC, Landsat 5, Remote sensing, Satellite imagery, SEBAL.