t.rast.wetsnow detects wet snow in dBi-scaled
backscatter data from Sentinel-1 C-band SAR. It is based on an algorithm
by Nagler et al. (2016) for wet snow detection (https://doi.org/10.3390/rs8040348).
input is a time series of geocoded Sentinel-1 C-band SAR backscatter
data scaled to dbi values, where polarization and
track information are part of the semantic labels of the
maps in the input SpaceTimeRasterDataset (STRDS).
Time steps are skipped when either the VV or VH input raster contains no
backscatter values at or above nodata_threshold. All backscatter pixels
with values below the threshold are considered as no data.
output is a SpaceTimeRasterDataset (STRDS) where wet snow detection results
will be registered. It will be created or extended if the e-flag is set.
Resulting raster maps within the STRDS contain a binary classification where the
value 1 represents "wet snow" and 0 represents "no wet snow". All input pixels
considered as no data are also marked as no data in the output.
Reference data are expected to be provided as imagery group(s) for each
track that can be found using the reference_pattern.
The pattern may include the mapset name where the reference data is stored.
The group-name pattern must contain exactly one * wildcard marking
the portion containing the track number, for example
Sentinel_1_reference_*. Other wildcard syntax such as ?
or character classes is not supported. The wildcard-matched portion should
contain only one number, the one-to-three-digit track number. Avoid years,
dates, or other numbers in this portion: extraction uses the first sequence
of up to three digits. For example, Sentinel_1_reference_2026_110
matched by Sentinel_1_reference_* is interpreted as track
202, not 110. Use Sentinel_1_reference_2026_* instead.
If the reference_pattern includes the @mapset suffix, reference
data is read only from the specified mapset. Ensure that only one valid reference
group matches each track. If several do, a warning is issued and the last group
returned by g.list is used.
The raster maps in the reference data groups are expected to have the following
semantic labels:
VH: Average backscatter in VH polarization (required)
VV: Average backscatter in VV polarization (required)
linc: Local incidence angle (optional if linc_weight is present)
linc_weight: Weight for the local incidence angle (optional)
mask_high_resolution: High-resolution binary mask indicating valid pixels
(combination of landcover and layover shadow) (required)
mask: Binary mask resampled to target resolution (optional)
Malnes, E. 2024: Deteksjon av våt snø med Sentinel-1. Teknisk Notat NORCE (unpublished)
Notat nr. 1. 105910
Malnes, E. & Guneriussen, T. 2002: Mapping of snow covered area with Radarsat in Norway.
IEEE International Geoscience and Remote Sensing Symposium (1):683-685.
Nagler, T. & Rott, H. 2000: Retrieval of wet snow by means of multitemporal SAR data.
IEEE Transactions on Geoscience and Remote Sensing. (38):754-765. 10.1109/36.842004.
Nagler, T., Rott, H., Ripper, E., Bippus, G., & Hetzenecker, M. 2016: Advancements
for snowmelt monitoring by means of Sentinel-1 SAR. Remote Sensing, 8(4):348.
Vickers, H., Malnes, E., & Eckerstorfer, M. (2022). A Synthetic Aperture Radar Based
Method for Long Term Monitoring of Seasonal Snowmelt and Wintertime Rain-On-Snow Events
in Svalbard. Frontiers in Earth Science (10):868945.