HOW DO AMPLITUDE AND PHASE SHIFT OF DIURNAL DISCHARGE FLUCTUATIONS AFFECT STREAM TEMPERATURE MODELS?

The use of energy balance models for predicting stream temperature through time and space often focuses heavily on characterizing heat fluxes, while simplifying the hydrologic system by assuming constant discharge through time. However, some streams exhibit diurnal discharge fluctuations due to evapotranspiration, diurnal glacial melt water inputs, and/or anthropogenic processes, with daily amplitude variations up to 50%. In certain situations, ignoring discharge fluctuations may yield large errors in the modeled stream temperature. In order to gain a better understanding of how the amplitude and timing of diurnal discharge fluctuations affect modeled stream temperatures, we updated the HFLUX stream temperature solver, a MATLAB based one-dimensional finite difference model, to allow for variable stream discharge through time. A scaling relationship between discharge and stream dimensions was developed based on Manning’s equation and a triangular channel shape, which allows for model node volume and stream width to automatically change through time as a function of discharge. The feedback relationships between discharge, water volume, width, and heat flux complicates predictions of how diurnal discharge fluctuations will affect stream temperatures. Field data, from a proglacial stream in the Peruvian Andes that experiences diurnal discharge fluctuations due to glacial melt water, were used in a case study. Model inputs include in-stream temperature sensor data along an approximately 1.2 km reach recorded at 5 minute intervals over about 5 days, climate data recorded at 10 minute intervals, streambed temperature recorded at 5 minute intervals, discharge recorded at 15 minute intervals, groundwater temperature, shading, cloudiness, stream dimensions, and sediment type. Model outputs from simulations using a range of amplitudes and phase shifts of the diurnal discharge were compared. Comparison of these simulations reveals which amplitude and timing combinations cause the largest effect on modeled stream temperatures. Our results demonstrate that incorporating diurnal discharge fluctuations can be important for modeling stream temperatures under certain circumstances.