Cold Air Outbreaks
Description
Overview: This study uses a set of criteria to examine cold air outbreaks (CAOs) across the globe from 1979 – 2018 and to determine how CAOs have changed over the last 40 years. We found CAOs occur most frequently in the Northern Hemisphere, with as many as 8 CAO days per year in North America and Eurasia. CAOs were found to have decreased in size, intensity, frequency, and duration across much of the globe, with the largest decreases in Alaska, Canada, and the North Atlantic, while an increase in CAOs was observed in Eastern Europe, Central Eurasia, and the Southern Ocean. Early and late winter CAOs have also become much less frequent in most regions. Data Used: Two-meter temperature (T2m) data was acquired from the NCEP/NCAR (NNR) climate reanalysis dataset (National Center for Environmental Prediction (NCEP) and National Center for Atmospheric Research (NCAR) and the recently released ERA5 reanalysis data set from the European Center for Medium-Range Weather Forecasts (ECMWF). ERA5 T2m was acquired at a 1 degree spatial resolution on an hourly timescale and converted to daily mean T2m while NNR daily mean T2m was acquired at a T62 gaussian grid (192 longitude and 94 latitude) spatial resolution from 1979 - 2018. CAO Methods: Three criteria for a CAO were designed to capture the most extreme CAOs while being flexible enough to capture the entire evolution of the event. 1.) Magnitude: The magnitude criterion requires the daily mean temperature to be at or below the 2.5th percentile threshold of deseasonalized 2-meter temperature (T2m). The daily mean T2m must also be at or below 20 degrees Celsius with a departure from the climatological mean of at least -2 degrees Celsius. 2.) Spatial Extent: The daily spatial extent, which is a summation of all contiguous grid points that meet the magnitude criteria, must be at least 1,000,000 km2. 3.) Duration: The duration criterion requires the magnitude criterion for the entire CAO be met for at least five consecutive days and begins on the first day in which the spatial extent criterion is met and ends on the last day the spatial extent criterion is met. How to use and interpret data: There are 3 files: 1.) and excel file of all CAOs for both the NNR and ERA5 (separate tabs). Because the ERA5 data is the primary data set used in this study it has two additional columns of data, one for the region of the CAO and one for the hemisphere of the CAO. 2.) A .mat file (MATLAB) of all the ERA5 CAO data. The column headers are as follows: [1. daily data for each CAO event, 2. onset date, 3. duration, 4. Mean z-score 5. mean z-score per gridpoint, 6. total duration per gridpoint 7. daily z-score per gridpoint 8. temperature anomaly each day, 9. Region 10. hemisphere] 3.) A similar .mat file, but for the NNR CAOs. Differences: columns 4 and 5 and 11 in the NNR file are not in the ERA5 file (shift headers). These were used in calculations but omitted from ERA5 file for size restraints.
Files
Steps to reproduce
Will need a MATLAB license.