Snow Telemetry Data Research Articles

Changes in Climate and Land Cover Affect Seasonal Streamflow Forecasts in the Rio Grande Headwaters

AbstractSeasonal streamflow forecast bias, changes in climate, snowpack, and land cover, and the effects of these changes on relations between basin‐wide snowpack, SNOw TELemetry (SNOTEL) station snowpack, and seasonal streamflow were evaluated in the headwaters of the Rio Grande, Colorado. Results indicate that shifts in the seasonality of precipitation and changing climatology are consistent with periods of overprediction and underprediction in streamflow forecasts. Multiple linear regression of SNOTEL data, postcedent precipitation, and land‐cover changes explained 2%–18% more variability in streamflow prediction than using SNOTEL station data alone. Simulated basin‐wide snowpack from a physically based model had significant negative trends in snow water equivalent (−4.33 mm/yr) and snow‐covered area (−0.05%/yr) during the melt period April–June. Simulated streamflow from a precipitation‐runoff model increased an average 5% when the effects of bark beetle‐induced tree mortality were compared to a baseline simulation with static vegetation. The effects of a 2013 wildfire increased simulated seasonal streamflow an average 35% for 1–4 years postfire. The combined effects of climate and land‐cover changes on snowpack‐streamflow relations highlight the difficulty in seasonal streamflow forecasting, which has important implications for water‐resource management.

Patterns of trends in niveograph characteristics across the western United States from snow telemetry data

The snowpack is changing across the globe, as the climate warms and changes. We used daily snow water equivalent (SWE) niveograph (time series of SWE) data from 458 snow telemetry (SNOTEL) stations for the period 1982 through 2012. Nineteen indices based on amount, timing, time length, and rates were used to describe the annual temporal evolution in SWE accumulation and ablation. The trends in these annual indices were computed over the time period for each station using the Theil-Sen slope. These trends were then clustered into four groups to determine the spatial pattern of SWE trends. Temperature and precipitation data were extracted from the PRISM data set, due to the shorter time period of temperature measurement at the SNOTEL stations. Results show that SNOTEL stations can be clustered in four clusters according to the observed trends in snow indices. Cluster 1 stations are mostly located in the Eastern- and South-eastern most parts of the study area and they exhibit a generalized decrease in the indices related with peak SWE and snow accumulation. Those stations recorded a negative trend in precipitation and an increase in temperature. Cluster 4 that is mostly restricted to the North and North-west of the study area shows an almost opposite pattern to cluster 1, due to months with positive trends and a more moderate increase of temperature. Stations grouped in clusters 2 and 3 appear mixed with clusters 1 and 4, in general they show very little trends in the snow indices.

Developing an intelligent expert system for streamflow prediction, integrated in a dynamic decision support system for managing multiple reservoirs: A case study

Extracting and using of time-dependent indices improved prediction accuracy.Pre-processing of data improved prediction accuracy.Intelligent selection of predictors via sensitivity analysis and data mining.Successful integration of a novel forecast expert system in an operation system. Since fresh water is limited while agricultural and human water demands are continuously increasing, optimal prediction and management of streamflows as a source of fresh water is crucially important. This study investigates and demonstrates how data preprocessing and data mining techniques would improve the accuracy of streamflow predictive models. Based on easily accessible Snow Telemetry data (SNOTEL), four streamflow prediction models autoregressive integrated moving average (ARIMA), artificial neural networks (ANNs), a hybrid-model of ANN and ARIMA (ANN-ARIMA), and an adaptive neuro fuzzy inference system (ANFIS) were developed and utilized in a streamflow prediction process on Elephant Butte Reservoir. Utilizing the statistical correlation analysis and the extracting importance degrees of predictors led to efficiently select the most effective predictors for daily and monthly streamflow to Elephant Butte Reservoir. For the daily prediction time step, by preprocessing the historical data and extracting and utilizing the extracted climate variability indices through data mining techniques, the ANFIS model achieved a superior streamflow prediction performance for Elephant Butte Reservoir compared to the other three evaluated prediction models. Additionally, for predicting monthly streamflow to the Elephant Butte Reservoir, ANFIS showed significantly higher accuracy than the ANNs. As an optimal application of the developed predictive expert systems, successful integrating the prediction models in integrated reservoir operations balanced the need for a reliable supply of irrigation water against losses through evaporation. The optimal operation plan significantly minimizes the total evaporation loss from both reservoirs by providing the optimal storage levels in both reservoirs. This study provides the conceptual procedures of non-seasonal (ARIMA) model, and since the model is univariate, it demonstrates a strongly-reliable inflow prediction when existing information is limited to streamflow data as a predictor.

Deriving snow-cover depletion curves for different spatial scales from remote sensing and snow telemetry data

During the melting of a snowpack, snow water equivalent (SWE) can be correlated to snow-covered area (SCA) once snow-free areas appear, which is when SCA begins to decrease below 100%. This amount of SWE is called the threshold SWE. Daily SWE data from snow telemetry stations were related to SCA derived from moderate-resolution imaging spectroradiometer images to produce snow-cover depletion curves. The snow depletion curves were created for an 80 000 km2 domain across southern Wyoming and northern Colorado encompassing 54 snow telemetry stations. Eight yearly snow depletion curves were compared, and it is shown that the slope of each is a function of the amount of snow received. Snow-cover depletion curves were also derived for all the individual stations, for which the threshold SWE could be estimated from peak SWE and the topography around each station. A station's peak SWE was much more important than the main topographic variables that included location, elevation, slope, and modelled clear sky solar radiation. The threshold SWE mostly illustrated inter-annual consistency. Copyright © 2015 John Wiley & Sons, Ltd.

A sensitivity study of high-resolution regional climate simulations to three land surface models over the western United States

The Weather Research Forecasting model is applied for convection-permitting regional climate simulations over the western United States using three different land surface schemes (Noah, NoahMP, and CLM). Simulated precipitation, temperature, and snow water equivalent (SWE) are evaluated by comparing against Snow Telemetry (SNOTEL) and Parameter-elevation Regressions on Independent Slopes Model (PRISM) observations. The results show that all simulations realistically reproduce the spatial and temporal variability of precipitation without significant sensitivity to the choice of land surface scheme, even though they tend to overestimate the magnitude of the SNOTEL data by about 15%. Comparing the bias with respect to the SNOTEL data, CLM is superior in 2 m maximum temperature, while NoahMP is most skillful in 2 m minimum temperature. Land surface parameterizations have high impacts on snowpack simulations. The SWE peaks too early with an unrealistically low value and also ablates too fast in Noah. NoahMP improves the SWE estimate to some extent, and CLM best represents the observations. Overall, CLM and NoahMP outperform Noah. Further analysis reveals that these differences are largely attributed to distinct rainfall-snowfall partitioning, snow albedo treatment, vegetation treatment, and surface data in these schemes.

Capitalizing on the daily time step of snow telemetry data to model the snowmelt components of the hydrograph for small watersheds

AbstractDaily time step data from National Resources Conservation Service snow telemetry (SNOTEL) stations were used to estimate components of the snowmelt‐dominated hydrograph for three small watersheds in Colorado and Wyoming. Thirty‐three years of SNOTEL data was paired with streamflow data to estimate the annual run‐off volume (Q100) and peak streamflow (Qpeak), as well as the timing of 20, 50, and 80% (tQ20, tQ50, and tQ80) of the annual streamflow. Multivariate regression models were derived from snow water equivalent (SWE) data for one basin (Joe Wright Creek), and the model components were then retained and applied to the other two watersheds. The basin‐specific coefficients developed for the Joe Wright model were also applied to the other two basins to evaluate direct transferability.Peak SWE, date of peak SWE, and number of consecutive days with snow on the ground up to the date of peak SWE had the strongest correlation to streamflow and formed the basis of the models. The optimal model configuration performed well for the tQ20, tQ50, and Q100, with poorer performance for Qpeak and tQ80. Applying the Joe Wright Creek coefficients directly to the other watersheds yielded reasonable results. Copyright © 2014 John Wiley & Sons, Ltd.

A Bayesian hierarchical model for forecasting intermountain snow dynamics

Because of a continual increase in the demand for water as well as an ongoing regional drought, there is an imminent need to monitor and forecast water resources in the Western United States. In particular, water resources in the Intermountain West rely heavily on snow water storage. Thus, the need to improve seasonal forecasts of snowpack and considering new techniques would allow water resources to be more effectively managed throughout the entire water‐year. Many available models used in forecasting snow water equivalent (SWE) measurements require delicate calibrations. In contrast to the physical SWE models most commonly used for forecasting, we offer a statistical model. We present a data‐based statistical model that characterizes seasonal SWE in terms of a nested time series, with the large scale focusing on the inter‐annual periodicity of dominant signals and the small scale accommodating seasonal noise and autocorrelation. This model provides a framework for independently estimating mainly the temporal dynamics of SWE for the various snow telemetry sites. We use snow telemetry data from 10 stations in Utah over 34 water‐years to implement and validate this model. Copyright © 2014 John Wiley & Sons, Ltd.

Evaluation of snowmelt simulation in the Weather Research and Forecasting model

The objective of this study is to better understand and improve snowmelt simulations in the advanced Weather Research and Forecasting (WRF) model by coupling it with the Community Land Model (CLM) Version 3.5. Both WRF and CLM are developed by the National Center for Atmospheric Research. The automated Snow Telemetry (SNOTEL) station data over the Columbia River Basin in the northwestern United States are used to evaluate snowmelt simulations generated with the coupled WRF‐CLM model. These SNOTEL data include snow water equivalent (SWE), precipitation, and temperature. The simulations cover the period of March through June 2002 and focus mostly on the snowmelt season. Initial results show that when compared to observations, WRF‐CLM significantly improves the simulations of SWE, which is underestimated when the release version of WRF is coupled with the Noah and Rapid Update Cycle (RUC) land surface schemes, in which snow physics is oversimplified. Further analysis shows that more realistic snow surface energy allocation in CLM is an important process that results in improved snowmelt simulations when compared to that in Noah and RUC. Additional simulations with WRF‐CLM at different horizontal spatial resolutions indicate that accurate description of topography is also vital to SWE simulations. WRF‐CLM at 10 km resolution produces the most realistic SWE simulations when compared to those produced with coarser spatial resolutions in which SWE is remarkably underestimated. The coupled WRF‐CLM provides an important tool for research and forecasts in weather, climate, and water resources at regional scales.

Trends in snowpack chemistry and comparison to National Atmospheric Deposition Program results for the Rocky Mountains, US, 1993–2004

Seasonal snowpack chemistry data from the Rocky Mountain region of the US was examined to identify long-term trends in concentration and chemical deposition in snow and in snow-water equivalent. For the period 1993–2004, comparisons of trends were made between 54 Rocky Mountain Snowpack sites and 16 National Atmospheric Deposition Program wetfall sites located nearby in the region. The region was divided into three subregions: Northern, Central, and Southern. A non-parametric correlation method known as the Regional Kendall Test was used. This technique collectively computed the slope, direction, and probability of trend for several sites at once in each of the Northern, Central, and Southern Rockies subregions. Seasonal Kendall tests were used to evaluate trends at individual sites. Significant trends occurred during the period in wetfall and snowpack concentrations and deposition, and in precipitation. For the comparison, trends in concentrations of ammonium, nitrate, and sulfate for the two networks were in fair agreement. In several cases, increases in ammonium and nitrate concentrations, and decreases in sulfate concentrations for both wetfall and snowpack were consistent in the three subregions. However, deposition patterns between wetfall and snowpack more often were opposite, particularly for ammonium and nitrate. Decreases in ammonium and nitrate deposition in wetfall in the central and southern rockies subregions mostly were moderately significant ( p<0.11) in constrast to highly significant increases in snowpack ( p<0.02). These opposite trends likely are explained by different rates of declining precipitation during the recent drought (1999–2004) and increasing concentration. Furthermore, dry deposition was an important factor in total deposition of nitrogen in the region. Sulfate deposition decreased with moderate to high significance in all three subregions in both wetfall and snowpack. Precipitation trends consistently were downward and significant for wetfall, snowpack, and snow-telemetry data for the central and southern rockies subregions ( p<0.03), while no trends were noted for the Northern Rockies subregion.

Snow water equivalent interpolation for the Colorado River Basin from snow telemetry (SNOTEL) data

Inverse weighted distance and regression nonexact techniques were evaluated for interpolating methods snow water equivalent (SWE) across the entire Colorado River Basin of the western United States. A 1‐km spacing was used for the gridding of snow telemetry (SNOTEL) measurements for the years 1993, 1998, and 1999, which on average, represented higher than average, average, and lower than average snow years. Because of the terrain effects, the regression techniques (hypsometric elevation and multivariate physiographic parameter) were found to be superior to the weighted distance approaches (inverse distance weighting squared, and optimal power inverse distance weighting). A regression detrended inverse weighted distance method was developed for the hypsometric and multivariate techniques in order to preserve the point SNOTEL data. On the basis of root mean square error analysis and estimates of SWE volumes in different elevation zones for the entire basin and for subbasins the elevation detrended method with a point by point regression was found to be the most appropriate technique. Various search radii and anisotropies of the search ellipse were tested with the hypsometric method, producing only small difference in the root mean square error and SWE volumes.