A New Snow Density Parameterization for Land Data Initialization

Abstract

Snow initialization is crucial for weather and seasonal prediction, but the National Centers for Environmental Prediction (NCEP) operational models have been found to produce too little snow water equivalent, partly because they assume a constant and unrealistically low snow density for the snowpack. One possible solution is to use the snow density formulation from the Noah land model used in NCEP operational forecast models. While this solution is better than the constant density assumption, the seasonal evolution of snow density in Noah is still found to be unrealistic, through the evaluation of both the offline Noah model output and the Noah snow density formulation itself. A physically based snow density parameterization is then developed, which performs considerably better than the Noah parameterization based on the measurements from the SNOTEL network over the western United States and Alaska. It also performs better than the snow density schemes used in three other models. This parameterization could be easily implemented in NCEP operational snow initialization. With the consideration of up to 10 snow layers, this parameterization can also be applied to multilayer snowpack initiation or to estimate snow water equivalent from in situ and airborne snow depth measurements.

The phenology of the subnivium

“…>0.25 g cm −3 ) maintenance periods decreased even with increasing snow depth, reducing subnivium maintenance from approximately 30 days at shallow depths (5-25 cm) to 10-15 days at midrange depths (25-45 cm, figure 5(b)). In fact, the range of snow densities at our surveyed sites (0.10-0.30 g cm −3 ) represents a low to moderate range compared to those found in other studies at similar latitudes (Dawson et al 2017, Derksen et al 2014.…”

Snowpack Change From 1982 to 2016 Over Conterminous United States

“…It is developed by consistently assimilating in situ measurements of SWE and/or SD at thousands of sites (Broxton, Dawson, & Zeng, ) and 4‐km gridded PRISM precipitation and temperature data (Daly et al, ) over ConUS. The details of the methodology and the robustness and accuracy of the data set have been reported (Broxton, Dawson, & Zeng, ; Broxton, Zeng, & Dawson, ; Dawson et al, , ). Here we outline key steps in generating this data set: The ratio of observed SWE over estimated net snowfall (accumulated snowfall minus accumulated snow ablation), rather than SWE itself, is used for interpolation from point measurements to other points or pixels (Broxton, Dawson, & Zeng, ). The snowfall versus rainfall is separated using a daily 2‐m air temperature threshold based on station data, and the snow ablation is also estimated as a function of temperature based on station data (Broxton, Dawson, & Zeng, ). A new snow density parameterization (Dawson et al, ) is developed to combine the SWE and SD measurements from hundreds of SNOTEL sites with the SD measurements from thousands of COOP sites. These steps are combined with the PRISM gridded daily temperature and precipitation data (Daly et al, ) to produce the UA daily 4‐km SWE and SD (Broxton, Dawson, & Zeng, ). SCE is computed from UA SWE with a threshold value of 3 mm. …”

A Wet‐Bulb Temperature‐Based Rain‐Snow Partitioning Scheme Improves Snowpack Prediction Over the Drier Western United States

“…Similar to many other LSMs that use T a for determining the precipitation phase (Pan et al, 2003;Toure et al, 2016), TDRY produces less snow in terms of snow depth ( Figure 2c) and SWE (Figure 2d) by −23.3% and −20.7%, respectively, over the drier Western CONUS. The relative biases of the modeled SWE are not consistent with those of snow depth; this may result from the slightly different snow density schemes used in Noah-MP (Niu et al, 2011) and the UA snow product (Dawson et al, 2017). Compared to the drier Western CONUS, TDRY performs better over the humid Eastern CONUS, resulting in a lower relative bias in snow depth and SWE of −3.1% and −3.3%, respectively, with a mixture of positive and negative biases (Figures 2c to 2d).…”

“…Compared to the drier Western CONUS, TDRY performs better over the humid Eastern CONUS, resulting in a lower relative bias in snow depth and SWE of −3.1% and −3.3%, respectively, with a mixture of positive and negative biases (Figures 2c to 2d). The relative biases of the modeled SWE are not consistent with those of snow depth; this may result from the slightly different snow density schemes used in Noah-MP (Niu et al, 2011) and the UA snow product (Dawson et al, 2017). Over the entire CONUS, TDRY produces less snow by −16.5% and −16.8% in terms of snow depth and SWE due mainly to the significantly low values over the mountain ranges in the Western CONUS.…”