Example gallery#
One short script per product: load it over one area of interest, say which routes serve it, and draw one or two figures. Every script is executed when the docs are built, so a figure on this page is a figure the package produced against live data on that day.
The sections follow the package: esd.stations, esd.snow, esd.sar,
esd.optical, esd.terrain, esd.land, esd.hydro, esd.boundaries, and
esd.climate, then the cross-cutting tools. The stations section is followed
by Other station datasets: the clients-only station data that is kept apart
from the five daily networks. Pushes to main and the weekly
scheduled build run every script with the provider credentials
(Credentials); a pull-request build runs only the scripts that need no
account and reuses the last output of the rest.
Each page offers the script and a Jupyter notebook to download.
Stations#
Point observations from five public snow networks: USDA NRCS AWDB (SNOTEL, SCAN, and snow courses), California’s CDEC, British Columbia’s DataBC, Norway’s NVE, and Yukon’s snow survey; the daily archive that pre-downloads all five; and one call that reads across every network. Every station here has a daily SWE or snow depth record; datasets that are mostly periodic or snow-depth only have their own section, Other station datasets, below.
Other station datasets#
Station data that is not part of the five daily networks above. These
datasets are reached only through their clients in
easysnowdata.stations.clients: they are not in esd.stations.inventory(),
esd.stations.load(), the catalog, or the pre-downloaded daily archive that
global_snow_networks publishes, and they are not mixed into the “all
networks” example.
They differ from the daily networks in what they measure and how often.
Periodic SWE. Manual snow courses and route surveys measured every one to four weeks: CanSWE and NorSWE (mostly), Finland’s SYKE, Switzerland’s SLF and Russia’s RIHMI-WDC.
Snow depth, daily. GHCN-Daily worldwide, DWD in Germany: many stations, but daily SWE at only a few. Finland’s FMI (automatic, climate and precipitation stations) publishes snow depth and no SWE. The Southern Andes dataset (Chile and Argentina) is snow depth only.
Daily SWE that is new. Hydro-Québec’s gamma sensors (a rolling 10-day feed), the automated stations in CanSWE, Alberta’s pillows (which AWDB also mirrors), Argentina’s snow pillows (SNIH) and CEAZA’s SWE sensors in the Chilean Andes.
Every record carries an interval ("daily", "periodic", "hourly",
…), so check it before treating a series as daily. The clients return plain
lists of dicts; the examples turn them into pandas and geopandas objects in a
few lines.
Alberta snow pillows and snow courses (Alberta River Basins)
Recent Russian snow surveys and snow cover (RIHMI АИСОРИ-М)
Southern Andes daily snow depth (Chile and Argentina)
Snow#
Snow cover from Terra and Aqua MODIS (daily NDSI snow cover, 8-day maximum snow extent, and the cloud-gap-filled daily product, 500 m) and from VIIRS on Suomi-NPP and NOAA-20 (daily and cloud-gap-filled, 375 m); snow water equivalent and snow depth from SNODAS (daily, 1 km, CONUS) and the UCLA snow reanalysis (daily, 480 m, western US and High Mountain Asia); the Sturm & Liston seasonal snow classes; and the Wrzesien mountain snow mask.
SAR#
Sentinel-1 C-band gamma0 backscatter (VV and VH), terrain-corrected, from Planetary Computer (10 m), OPERA RTC-S1 (30 m), or Earth Engine, and the local incidence angle with the layover and shadow mask, from the OPERA static layers or computed from any DEM.
Sentinel-1 RTC backscatter and local incidence angle
Optical imagery#
Surface reflectance from Sentinel-2 L2A (10–60 m, with the scene classification layer), HLS v2.0 (Landsat 8/9 and Sentinel-2 at 30 m on one grid, with Fmask), and PlanetScope (3 m, commercial, with the UDM2 usable-data mask).
Terrain#
Five digital elevation models behind one load (Copernicus GLO-30 and GLO-90,
NASADEM, SRTM GL1, USGS 3DEP, and ALOS World 3D), the CHILI heat-load index, and
Natural Earth shaded relief as a hillshade basemap.
Digital elevation models: Copernicus, NASADEM, SRTM, 3DEP, ALOS, GEDTM30
Hillshade: Natural Earth shaded relief as a basemap
Land cover#
ESA WorldCover (global, 10 m, 2020 and 2021), the US National Land Cover Database (30 m, annual 1985–2024 and the 2021 release), and the Copernicus CGLS-LC100 tree-cover fraction (global, 100 m, 2015–2019).
Hydrography#
Basin boundaries: the USGS Watershed Boundary Dataset (HUC2 to HUC16, United States), HydroBASINS and BasinATLAS nested sub-basins (global, Pfafstetter levels 1–12), the GRDC major river basins, and the GRDC / WMO basins and sub-basins.
Boundaries#
Countries, first-level subdivisions (states and provinces), US counties, and administrative units at any level (Natural Earth, the US Census Bureau, and geoBoundaries); Natural Earth map layers; GMBA mountain ranges; and RGI 7.0 and 6.0 glacier outlines.
Glacier outlines: Randolph Glacier Inventory 7.0 and 6.0
Natural Earth map layers: lakes, rivers, glaciers, places
Climate and reanalysis#
ERA5 hourly reanalysis (ARCO-ERA5 on Google Cloud, 0.25°, 1940 onward), ERA5 and ERA5-Land hourly, daily, and monthly aggregates on Earth Engine, and the Köppen-Geiger climate classes (Beck et al. 2023) for five historical periods and the CMIP6 projections, and the monthly climate-mode indices (PDO, RONI/ONI, AO, NAO, PNA, AMO, MEI.v2) from NOAA NCEI, CPC and PSL.
Climate-mode indices: PDO, ENSO, AO (NOAA NCEI / CPC / PSL)
Köppen-Geiger climate classification (Beck et al.)
Tools#
The helpers every product shares: areas of interest in any form, time ranges, water years and day-of-water-year coordinates, and the catalog itself.
