Overview of the Global Climatological Analysis Toolkit (GCAT)
The Global Climatological Analysis Tool (GCAT) developed by NCAR has been used in the last 15 years by the US Army to define typical weather and dosage patterns in any area of the world at the local scale. This workflow system deployed on the DoD HPC features the use of the Weather and Research Forecasting system (WRF) model that, with its Climate Four Dimensional Data Assimilation (C-FDDA) function to assimilate any observation within the domain, downscales to 1.1 or 3.3 km resolution over any location globally on demand. Post-processing features include machine learning- based downscaling, Self Organizing Maps (SOM) clustering and classification to define typical days, Second-order Closure Integrated Puff diffusion model (SCIPUFF) execution, and subseasonal to seasonal ensemble forecasting driven by a time-lagged ensemble of the National Center for Environmental Prediction (NCEP) Climate Forecast System version 2 (CFSv2). An overview and applications of the system will be presented along with other emerging GCAT capabilities such as low wind speed event SOM analysis, and the hybrid physics-based ensemble seasonal forecasting.
IMPACT
GCAT is a WRF workflow system which provides downscaled weather globally in user defined regions and time periods. Products include Seasonal forecasting, AI / ML downscaled subdomains, AI / ML reduced climatological datasets, historical reanalysis, and SCIPUFF functionality. The simulations are used for mission planning efforts which require planning windows on a short turn around time scale. The system has been in place for over a decade with typically yearly updates dictated by the user base and current technological advancements.
PRESENTER
meech, scott
smeech@ucar.edu
727-798-6912
NCAR
CO-AUTHOR(S)
stefano alessandrini
alessand@ucar.edu
CATEGORY
Weather &  Ocean Modeling & Sim
SECONDARY CATEGORY
AI/ML for HPC
SYSTEM(S) USED
Carpenter