WEATHER & CLIMATE MODELING
PROJECT // CHAMPS
While employed at the South Carolina Department of Natural Resources in the State Climate Office (SCO), Jason Caldwell - founder of Weather & Water developed an integrated modeling framework from conception to implementation for serving the emergency management and agricultural stakeholders throughout the State. At the time, the MM5 modeling software (now morphed into the Weather Research & Forecasting (WRF) model) was the state-of-the science numerical weather and climate prediction system for predicting short-, medium-, and long-range weather and climate scenarios for use in planning activities.
In collaboration with the local National Weather Service offices and Southeast River Forecast Center Hydrologist-in-Charge John Feldt at the time, the SCO aimed to build an infrastructure that would tie multiple models together - much like the modern National Water Model approach. Ahead of it's time, this is exactly the type of service Weather & Water seeks to provide today - leading edge data and model efforts.
PROJECT // MM5 QPF SENSITIVITY
Two major snowstorms hit eastern North Carolina during the year 2000: a blizzard dumping over 2 feet of snow in Raleigh in January and another storm in December 2000 aimed at the Research Triangle Park area with predictions of 12-24 inches of snowfall. Models that performed well in the January storm, performed poorly in the December storm with the highest snowfall occurring 100 miles east of Raleigh along the I-95 Corridor.
Working on the high-performance computing (HPC) clusters available through North Carolina State University, Jason (our founder) worked in collaboration with the National Weather Service Forecast Office, local television station WRAL, and Baron Services to design, implement, and examine a myriad of sensitivity experiments to discern the source of model errors in the once reliable ETA (now NAM) modeling system. From investigation into inital conditions (e.g., sea surface temperatures or SSTs) to various model physics and dynamic parameterization schemes, the final results indicated SSTs offshort were likely the cause as minimal differences resulted from the remainder of experiments. Weather & Water is dedicated to using these same skills to identify error sources, evaluate forecast performance, and implement solutions using high-resolution modeling approaches.