This dataset contains v3.1 of the Daily Snow Water Equivalent (SWE) product from the ESA Climate Change Initiative (CCI) Snow project, at 0.1 degree resolution.

Snow water equivalent (SWE) is the depth of liquid water that would result if the of snow cover melted completely, which equates to the snow cover mass per unit area. The SWE product covers the Northern Hemisphere from 1979/01 to 2022/05 with complex terrain, land ice, and large lakes masked. The dataset covers the Northern Hemisphere winter season (October – May; occasional data produced during June and September) at a daily frequency starting in October 1987 and every second day from 1979 to May 1987. Retrievals are not produced for coastal regions of Greenland.

The product combines passive microwave data with ground-based snow depth measurements, via Bayesian non-linear iterative assimilation, to estimate SWE. It is based on data from the recalibrated enhanced resolution CETB ESDR dataset (MEaSUREs Calibrated Enhanced-Resolution Passive Microwave Daily EASE-Grid 2.0 Brightness Temperature (CETB) Earth System Data Record (ESDR) https://nsidc.org/pmesdr/data/), resampled to the 12.5km EASE-Grid 2.0.

A background snow-depth field, derived from re-gridded snow-depth observations made at synoptic weather stations, and a passive microwave emission model are the key components of the retrieval scheme. Snow density, which varies in both time and space, is parameterized from interpolated in situ observations from snow courses and snow pillows equipped with co-located snow depth sensors.
The dataset is aimed to serve the needs of users working on climate research and monitoring activities, including the detection of variability and trends, climate modelling, and aspects of hydrology and meteorology.

The Finnish Meteorological Institute is responsible for the SWE product generation. The SWE development is carried out in collaboration by FMI and Environment and Climate Change Canada (ECCC).

Changes from v2.0 and v3.0
v3.1 applies spatially and temporally varying snow densities within the SWE retrieval instead of during post-processing. The dry snow detection algorithm as well as the snow masking in post-production have also been updated. The time series has been extended from snow_cci version 2 by two years from 2020 to 2022. In comparison with in situ snow courses, the correlation and RMSE of v3.1 improved by 0.014 and 0.6 mm, respectively, relative to v2.0. The timing of peak snow mass is shifted two weeks later compared to v1.0 and reduction in peak snow mass presented in v2.0 is removed in v3.1. Differences between v3.0 and v.3.1 are minor, the resampling from 12.5km EASE-Grid 2.0 to the final 0.1 resolution grid has been changed for v.3.1 resulting in improved peak snow mass estimation.