Meteorological Files – anywhere on the globe
Skills: Meteorological modelling, data processing, MMIF, WRF.
Katestone’s massive computing resources came in handy for a recent project to deliver a five year meteorological dataset using WRF grids down to a 600m resolution in a very short timeframe. The meteorological modelling was required to assess a refinery development located on a peninsular in Saudi Arabia.
We were commissioned by WardKarlson Consulting Group (WKC Group) on behalf of ARM Associates (ARM) to provide a meteorological dataset suitable for use in dispersion modelling, using CALPUFF, for a detailed industrial air quality assessment.
A quick turn-around for the generation of a three-dimensional five-year dataset, processed and ready to use directly with the CALPUFF model was crucial to the project.
Datasets can take a long time to generate. For example to run the WRF configuration for this project on a single high speed modelling computer (not including setup time, configuration testing or post processing) would take approximately 60 days. Katestone delivered the tested and verified dataset well within the clients timeframe of 20 days, utilising the parallel-computing capability of the WRF (Weather Research & Forecasting) model.
The Katestone High Performance Cluster (HPC), a network of processors and computers designed specifically for parallel computing, was optimised to run WRF, resulting in a significantly reduced model run time.
The Mesoscale Model InterFace (MMIF) tool was used to perform a direct pass-through of the WRF-generated dataset, without any adjustments or re-diagnosing of meteorological variables, bypassing CALMET and all its known problems.
A comparison of generated meteorological data with observations at three sites within the domain shows that the WRF model performed well, and produced a dataset suitable for a detailed assessment.
[thb_testimonials id=”8486″ items=”1″ style=”1″ title=”Testimonial”]