GWTC2 Dataset - A global-scale classification of air masses

Published: 3 February 2020| Version 2 | DOI: 10.17632/gbwwksnd6j.2
Cameron Lee


-v2 has been updated to fill missing data on the first day of each month between October 2018 and December 2019. The GWTC2 is the Gridded Weather Typing Classification - version 2 (global domain). The GWTC2 is a system (and dataset) for defining weather types or air masses - which can be thought of as multivariate 'weather situations' (e.g. hot and humid, or cold and dry) occurring near the surface. At every 0.5-degrees of latitude and longitude (259,920 grid-points), the GWTC2 categorizes every day, from 1979-2019 (and updated regularly) into one of 11 different air masses (AMs). These AMs are both geographically- and seasonally-relative, meaning, for example, that a Humid Warm air mass can occur at any place and at any time of year (even in the winter in the polar regions), and simply signifies that the weather on that day was warmer and more humid than 'normal' (for that time of the year, and for that location). While weather types at individual locations are mostly classified independently from nearby locations, when mapped, synoptic-scale air masses can be observed (see: There are a few other weather typing classifications available, but this is the first that is available at a fine-scale, gridded, and global domain, and the first available over maritime locations. Greater detail on the methodology and development of this dataset can be found here: Lee, C.C., 2019. The gridded weather typing classification version 2: A global‐scale expansion. International Journal of Climatology. DOI: 10.1002/JOC.6263. More information can also be found here:, along with updated historical data and forecasts. The compressed folder below contains a CSV-formatted file for each month, 1979-2019. In each CSV file, each row is a lat/long coordinate (indicated by the first two columns), and each column is an Excel-readable date (28856 = 01-Jan-1979). All other data in each file is the AM code for each location on each day. A metadata file has also been provided to help identify the 11 different air mass types. Note that missing values are indicated with NaN or 0.


Steps to reproduce

Steps to reproduce this data can be found in the referenced article, and citations therein.


Kent State University


Climatology, Climate Classification, Applied Climatology, Synoptic Climatology, Climate Change, Weather