Hiking up to the next plot…
This dataset contains maps of cold-air pooling (CAP) across Colorado at 100m resolution (Colorado_100m.nc) and the region surrounding the East River watershed, CO at 30m resolution (EastRiver_30m.nc). Classifications were determined using a Python implementation of the algorithm originally developed in Lundquist et al. (2008), applied to digital elevation models (DEMs) accessed and resampled from the U.S. Geologic Survey 3D Elevation Program (USGS, 2024) using the py3dep Python module (Chegini, 2021). This dataset is released in support of an upcoming manuscript studying CAP distributions and mountain weather station representativeness across these two regions (currently in preparation). The two NetCDF (.nc) files are structured as follows: Variables elevation: elevation at each pixel (meters). CAP: cold-air pooling signal at each pixel. Values: 1 = CAP, 0 = no-signal, -1 = no-CAP. spatial_ref: contains attributes with coordinate system metadata The EastRiver_30m.nc file additionally contains the three variables used to determine CAP signal: slope (degrees) rank: the fractional height of a pixel relative to its neighbors curvature: see the formula in Lundquist et al. (2008) or Cramblitt (2025) Coordinates x: easting (NAD83 / Conus Albers) y: northing (NAD83 / Conus Albers) xv_dist: cross-valley distance (meters). A critical input defining the critical radii r in rank and curvature calculations. Files Colorado_100m.nc: CAP classifications at 100m-resolution across the Colorado mountains for cross-valley distances from 10,000m to 250m (every 500m, with the addition of 250m). Users may select the value which best describes their region of interest. EastRiver_30m.nc: CAP classifications at 30m-resolution across terrain surrounding the East River watershed for a single cross-valley distance of 3,000m. For further detail on algorithm design and function, including the calculation of rank and curvature, CAP classification criteria, and choice of cross-valley distance, please refer to Lundquist et al. (2008) or the linked GitHub repository. Discussion on algorithm implementation and constraints will be released in an upcoming manuscript. Version History Version 2.1.0 removed the SNOTEL_CAP_classifications.csv file from the dataset. Version 2.0.0 updated the contents of all three files and renamed SNOTEL_CAP_classifications.csv to Colorado_SNOTEL_CAP.csv. The previous version was missing information near the state borders, which was corrected here by expanding the spatial domain in Colorado_100m.nc and adjusting the CAP classifications in Colorado_SNOTEL_CAP.csv accordingly. The updated files also reflect the recent corrections to the calculation of curvature and handling of coordinate systems in v2.0.0 of the CAP algorithm (Cramblitt, 2025). The effect of these corrections is relatively minor, but do affect the CAP signal at a few SNOTEL stations. Notably the coordinate reference system has been changed from WGS84 (EPSG:4326) to NAD83/Conus Albers (EPSG:5070). The variables 'slope', 'curvature', and 'rank' were removed from the Colorado_100m.nc file to decrease file size and allow for more manageable downloads. These variables are available upon request. Version 1.1.0 added the SNOTEL_CAP_classifications.csv file to the dataset. For questions, please contact jcramblitt@berkeley.edu.
Items connected by shared entities, co-authorship, citations, or semantic similarity.