Snow Water Equivalent Verification

The National Weather Service (NWS) is evolving quickly and new forecasting techniques are being implemented constantly. Forecasters at the local offices should be experts on mesoscale and sometimes microscale meteorological climates in their areas. Prior to the era of blending numerical weather prediction the biases from individual models were well known by forecasters and elements were adjusted accordingly to create a forecast. Unfortunately, the biases from model blends and national centers are unknown. This study is aimed at trying to figure out whether or not there are biases from model blended guidance. If so, how large are these biases, where are they occurring, how consistent are they, and what may be causing them.

In order to conduct this study snow water equivalent (SWE) from each SNOTEL site was recorded for each winter weather product issued by the Grand Junction forecast office during the 2019-2020 winter season. The SWE values were analyzed on an event basis and compared to the Quantitative Precipitation Forecasts (QPF) from the NWS, National Blend of Models (NBM), and Weather Prediction Center (WPC) during the same time period. SNOTELs were chosen for this study because a majority of our winter weather products are issued for the mountains, and there is an absence of people able to take manual observations in these often remote locations. Also, SWE amounts from these sites are used for our official verification. In addition to SWE other parameters like the flow regime were also documented in order to analyze if synoptic patterns impacted the biases.