Skip to content

Product Overview

Basic Information

The Cold Regions Snowfall and Snowpack Dataset (CoReSSD) is a 1 km snow water equivalent and precipitation dataset for North America. CoReSSD covers North America above 30°N for water years 2001-2021. This dataset was produced with funding from NASA Grant #80NSSC22K0585. CoReSSD relies on the Blender algorithm (Dechow et al., 2026, Water Resources Research). Blender is a non-linear optimization technique that recomputes the modeled mass and energy balance of the snowpack, constrained by remotely sensed snow cover fraction. Blender requires a prior estimate of the snowpack mass and energy balance.

CoReSSD Product Specifications

  • TEMPORAL COVERAGE: October 1, 2000 - September 30, 2021
  • TEMPORAL RESOLUTION: Daily
  • SPATIAL COVERAGE: North America above 30°N
  • SPATIAL RESOLUTION: 0.01° (nominally 1 km at equator)
  • DATA FIELD 1: Snow water equivalent [m]
  • DATA FIELD 2: Daily precipitation water-equivalent rate [m day-1]

The CoReSSD product is gridded at 0.01° (nominally 1 km at equator). The netCDF files have an equally spaced 0.01° grid for both latitude and longitude.

Input Data

Blender requires five input fields, plus an optional sixth field. The first four inputs are modeled components of the snowpack mass and energy balance, and the fifth is remotely sensed snow cover fraction (SCF). The optional sixth field, the precipitation scalar, is used when the SCF product is much higher resolution than the modeled fields or when mass inputs are not well distributed spatially. The second case applies to the production of the CoReSSD product.

Modeled Inputs

  • Snowf_tavg: snowfall rate [kg m-2 s-1]
  • SWE_tavg: snow water equivalent [kg m-2]
  • Qg_tavg: downward soil heat flux [W m-2]
  • Tair_f_tavg: air temperature [deg K]

The tavg field suffix indicates the field is a “time-average value” over the timestep. In the production of the CoReSSD prior, the timestep was daily.

Observed Inputs

  • SCF: pixel snow cover fraction [dimensionless]

Optional Inputs

  • precip_scalar: precipitation scalar value [dimensionless]

The precipitation scalar is an optional input to assist in downscaling the mass inputs. Further documentation for this input is available in Dechow et al., 2026 (still in review at Water Resources Research). If an input is not provided, the Blender algorithm fills a dummy value of 1, which effectively turns this option off.