The watersheds of the Northern Great Plains (NGP) have experienced a cycle of extreme wet and dry periods over the last four decades due to the highly variable cold and semiarid continental climate. The influences of such extreme climate shifts on the cold region hydrologic cycle result in devastating effects such as drought and flooding. Since the mid-1980s, watersheds in the NGP have faced a large and abrupt surge in precipitation with a few occasional short dry periods resulting in elevated streamflow and subsequent flooding. Nevertheless, the impacts of the recent wetting on snow processes such as snow accumulation, blowing snow transport, in-transit sublimation, fill and spill hydrology, frozen soil infiltration and streamflow generation mechanisms in headwater catchments are poorly understood.

In this study, we are utilizing a field-tested, physically-based and distributed cold region hydrologic model (CRHM) to detect underlying mechanisms of hydrologic changes to these climatic extremes in the Upper Sheyenne River Basin (USRB) above Harvey, ND during 2004-present. Snow surveys within the study area are being conducted to create a better understanding of snow processes. We have discretized the USRB into ~1000 hydrologic response units (HRU) based on land surface properties such as elevation, vegetation, slope, aspect, pothole density and its storage capacity. The model will be evaluated against distributed snow observation and streamflow. Our current model setup will decipher the underlying mechanism of hydrologic changes to recent wetting and explain the high variability of streamflow in the USRB.

Project Objectives:

The objective of this study is to detect the impacts of recent wetting on fill and spill hydrology, snow processes, and the genesis of streamflow (snowmelt, rainfall, rain on snow, contributing area, depressional storage change) by investigating intermediate processes, water balance, hydrologic state variable and land surface properties. The study area is currently experiencing a prolonged period of wetting that began in the early 1990’s, initially evidenced by an increase in the number of small water-filled depressions.

Progress:

Land surface evaluation

The Upper Sheyenne River Basin was discretized into ~1000 hydrologic response units on the basis of common physical characteristics, into areas less than 3 km². The study area for this project is the headwaters of the Sheyenne River (~1098.2 km² total drainage area; ~398.9 km² contributing drainage area), which is a headwater catchment and one of the major contributors to the Red River. The landscape of the entire basin gently slopes from southwest to northeast. Relief is low, approximately 158 m, in which the lowest elevation in the basin is ~ 471 m at the Harvey Dam spillway and the highest is ~ 629 m in the west-southwest of the drainage area. The area is generally flat, with low and rolling glacial hills. Major land covers are agricultural fields (~35-40%), herbaceous grasslands with sufficient depressional storage (~45%), pasture land and hay fields (~5-10%), and wetlands/open water (~10%).  The soil is primarily loam across all of the headwaters basin (USDA SSURGO). There are also small areas of sand, gravel, clay, and peat.

Snow sampling

Snow samples were collected within the watershed in the late winter of 2020 to quantify the water equivalent of the snowfall within the study area.

Model setup

The Cold Regions Hydrologic Modeling platform is used to compile a model for a watershed using a wide range of process structures where a module represents each process. In the headwaters of the Sheyenne basin, a model simulating hydrological processes needs to have modules representing prairie hydrologic responses. Thus, in this study, we will use following modules to represent the governing hydrological processes:

• Infiltration module: accounts for infiltration into frozen and unfrozen soils.
• Observation module: interpolates and adjusts the data, and allows modeled climate results to self-adjust to determine the impact of changes to other modules as a result.
• Radiation module: estimates the clear-sky direct and diffuse solar radiation as a function of latitude, elevation, ground slope, and azimuth, and accounts for overcast conditions based on incoming shortwave radiation (estimated by the Annandale module).
• Annandale module: estimates atmospheric transmittance from daily minimum and maximum temperatures using an empirical method.
• Longwave radiation module: estimates incoming longwave radiation using temperature, humidity, and atmospheric short-wave transmittance.
• Albedo module: calculates snow albedo through the winter and melt period, and the non-snow albedo after snow cover depletion.
• Canopy module: models the amount snowfall and rainfall intercepted by canopy; subsequent sublimation or release of intercepted snow by unloading or drip; and evaporation or drip of rainfall from the canopy.
• Blowing snow module: calculates snow transport and in-transit sublimation across hydrologic representative units (HRUs) using wind speed, air temperature, and relative humidity from the observation module.
• Energy balance snowmelt module: simulates snowmelt using the net energy balance by estimating shortwave and longwave radiation and other convective energy using semi-empirical techniques.
• All-wave radiation module: estimates net radiation from shortwave radiation for the period of the year during which the ground is not snow-covered in order to calculate evapotranspiration. 
• Evaporation module: uses the Penman-Monteith combination method to simulate evapotranspiration using surface resistance and available energy for evapotranspiration. Surface resistance depends on plant life-cycle status and growth, vapor pressure deficit, soil moisture availability, and the temperature at which transpiration can occur.
• Soil module: estimates moisture balance, depressional storage, overland flow, and subsurface flow in the soil and groundwater.
• Routing: utilizes the Muskingum routing method to transport from HRU to basin outlet.

Significance:

We anticipate that the long, slow observed increase in precipitation impacts snow accumulation, timing of peak SWE and snowmelt, duration of snow cover period, snowmelt runoff, streamflow generation mechanism, frozen vs. unfrozen soil conditions, and infiltration capacity. The use of a physically-based model in the CRHM platform will help to understand the impacts of this climatic shift on hydrologic processes in the Northern Great Plains over decades.

Recent outside research has found significant changes to the yearly/seasonal hydrologic cycle caused by the previous year’s climate; specifically, the amount of moisture frozen into the soil over the winter, the amount of snow received by the area over the winter, and the rate of the spring thaw. By expanding the study area to include watersheds that flow into larger bodies of water (in this case, as the Sheyenne River flows into the Red River of the North), the regional effects of seasonal patterns can be predicted by the model. Adding and refining the land-use component allows us to understand the impact that human activity could have on the watershed; for example, using traditional tillage versus no-till crop farming, or draining low-lying wetland areas. This contributes to our ability to prepare for floods and other climatic disasters, before they occur, on the basis of the previous year’s precipitation.

Conference/Seminar Presentations:

     Holmes, Stevie; Mahmood, Taufique; Mann, Michael. 2020. “Cold Region Hydrologic Variability to Recent Wetting in the Northern Great Plains.” Presented to the American Geophysical Union International Meeting, 1-19 December 2020, San Francisco, CA.