3S-GEOPROF-COMB: A Global Gridded Dataset for Cloud Vertical Structure from combined CloudSat and CALIPSO observations

Global cloud dataset from combined spaceborne radar and lidar.

This repository contains the 3S-GEOPROF-COMB product, a globally-gridded dataset for cloud vertical structure retrieved from hybrid active remote sensing (CloudSat radar and CALIPSO lidar) reported at 240 m vertical resolution. Science variables include vertical cloud fraction and vertically-integrated cloud cover for various geometrical criteria (i.e. high, middle, low, and thick clouds, along with with unique high, middle, and low cloud cover variants).

A Python notebook showing how to work with the dataset is available on GitHub, as is the source code used to produce the data product.

Our product is calculated from the latest release (R05) of per-orbit (level 2) combined cloud mask profiles in 2B-GEOPROF-LIDAR with additional data from 2B-GEOPROF. Validation and a complete description of the data product is given in the paper "A Global Gridded Dataset for Cloud Vertical Structure from Combined CloudSat and CALIPSO Observations" (Earth System Science Data).

Please cite "Bertrand, L., Kay, J. E., Haynes, J., and de Boer, G.: A global gridded dataset for cloud vertical structure from combined CloudSat and CALIPSO observations, Earth Syst. Sci. Data, 16, 1301–1316, https://doi.org/10.5194/essd-16-1301-2024, 2024."

The files contained in each folder are given via the following format:

          instruments_frequency_resolution.zip

• instruments:
  • radarlidar: the standard product, computed from merged geometrical profiles of hydrometeor occurrence
  • radaronly: computed solely from CloudSat radar profiles, otherwise processing is identical. For when users need to determine which instrument is responsible for observations of interest.
  • lidaronly: computed solely from CALIPSO lidar profiles, otherwise processing is identical. For when users need to determine which instrument is responsible for observations of interest.
• frequency:
  • monthly: data files report fields aggregated over a 1-month period
  • seasonal: data files report fields aggregated over a 3-month period (DJF, MAM, JJA, SON)
• resolution:
  • 2.5x2.5: each grid box spans 2.5 degrees latitude and 2.5 degrees longitude
  • 5x5: each grid box spans 5 degrees latitude and 5 degrees longitude
  • 10x10: each grid box spans 10 degrees latitude and 10 degrees longitude

Each folder contains a netCDF data file and a cloud cover quicklook plot image file for each time period over the 2006-2019 data record. Individual files are named according to the following format:

          timeperiod_instruments_datastream_version.nc (or .png)

• timeperiod: the time step at the given frequency, either e.g. 2006-08 (August 2006) or 2012-DJF (December 2012 to February 2013).
• instruments: the instruments used in the data product as a whole, always CSCAL (CloudSat and CALIPSO).
• datastream: either 3S-GEOPROF-COMB (COMBined radar and lidar), 3S-GEOPROF-COMB-RO (the auxiliary Radar Only variant of the product), or 3S-GEOPROF-COMB-LO (the auxiliary Lidar Only variant of the product)
• version: current release is v8.4

The product handles the 2011 CloudSat battery anomaly, after which the satellite only collects data in the sunlit portion of its orbit, by allowing users to subsample the pre-anomaly period to mimic the post-anomaly collection patterns. This allows users to estimate the effect of the reduced sampling on their analyses or apply a consistent sampling mode to the entire dataset. This option is provided to users via the "doop" dimension. Dimension coordinate value "All cases" reports variables computed using all observations, while "DO-OP observable" reports variables using only input data that either were or would have been collected in DO-OP mode (i.e. the pre-DO-OP period is subsampled to DO-OP collection patterns).