Customization of a Mesoscale Numerical Weather Prediction System

Customization of a Mesoscale Numerical Weather Prediction System for Energy & Utility Applications Anthony P. Praino and Lloyd A. Treinish Deep Computing Institute IBM Thomas J. Watson Research Center Yorktown Heights, NY, USA {lloydt, apraino}@us. ibm. com http: //www. research. ibm. com/weather/DT. html

Customization of a Mesoscale Numerical Weather Prediction System for Energy Industry Applications Ÿ Background and motivation Ÿ Architecture and implementation Ÿ Customization for energy applications –Energy Distribution –Energy Generation Ÿ Discussion, conclusions and future work

Background and Motivation Ÿ Estimated impact of weather on all types of energy & sanitary service across all geographic and temporal scales in the US is ~$230 B/year –$ 0. 1 B to $1 B per year for US energy industry related to poor temperature forecasts Ÿ Weather-sensitive utility operations are often reactive to short-term (3 to 36 hours), local conditions (city, county, state) due to unavailability of appropriate predicted data at this scale Ÿ Mesoscale (cloud-scale) NWP has shown "promise" for years as a potential enabler of proactive decision making for both economic and societal value

Background and Motivation Ÿ Despite the "promise" of cloud-scale NWP –Can models be coupled to weather sensitive business problems to demonstrate real value? –Can a practical and usable system be implemented at reasonable cost? Ÿ Evaluate concept via implementations in several location around the country. Ÿ New York domain has the longest operational history –Operational end-to-end infrastructure and automation with focus on HPC, visualization and system integration –Forecasts to 1 km resolution for metropolitan area with 3 to 21 hours lead time –Prototype applications with actual end users

Model Forecast Domains Ÿ Triply nested telescoping grids Ÿ Modelling code derived from highly modified version of non-hydrostatic RAMS Ÿ Explicit, full cloud microphysics Ÿ Typically, one or two 24 -hour runs per day Ÿ NAM-212/215 via NOAAport for lateral boundaries nudged every 3 hours Ÿ NAM-212/215 for initial conditions after isentropic analysis

Implementation and Architecture Ÿ Sufficiently fast (>10 x real-time), robust, reliable and affordable –E. g. , 1. 5 hours (42 x 375 MHz Power 3), 2. 0 hours (24 x 375 MHz Power 3) –Focus on HPC, visualization, system integration and automation Ÿ Ability to provide usable products in a timely manner Ÿ Visualization integrated into all components Weather Data Pre-processing Synoptic Model Processing Post-processing and Tracking RS/6000 SP ETA NOAAPORT Data Ingest FCST Boundary Conditions Other Input Products Advanced Visualization Weather Server NCEP Forecast Products Satellite Images Other NWS Data http: //www. . . Analysis Initial Conditions Forecast Modelling Systems Analysis Observations Data Explorer Data Assimilation Cloud-Scale Model Custom Products for Business Applications and Traditional Weather Graphics

Visualization Component Ÿ Traditional meteorological visualization is typically driven by data for analysis -- inappropriate for energy utility applications Ÿ Timely usability of cloud-scale NWP results requires –Understanding of how weather data need to be used for end users –Identification of user goals, which are mapped to visualization tasks –Mapping of data to visualization tasks –Users have limited control over content (targeted design) and simple interaction –Products designed in terms relevant for user Ÿ Wide range of generic capabilities needed –Line plots to 2 d maps to 3 d animations -- but customized –Assessment, decision support, analysis and communications –Automated (parallelized) generation of products for web dissemination –Highly interactive applications on workstations

Example Customizations for Utility Operations Ÿ Distribution operations Ÿ Generation operations

Electricity Transmission Ÿ New York State Transmission System –Color-contoured to show forecasted temperature –Available in 10 minute intervals from each 24 hour Deep Thunder forecast at 4 and 1 km resolution –Can be used to estimate transmission efficiency – 115 k. V and above Ÿ Map also shows –State and county boundaries –Major cities

Example -- Electricity Demand Forecasting Ÿ Simple estimated load –f(t, T, H) -color and height –Scaled by capacity –Generator data from Georgia Power –Deep Thunder forecast Ÿ Map shows –Heat index –State & county boundaries –Major cities –Generating plants

Emergency Planning for Severe Winds Ÿ Geographic correlation of demographic and forecast data Ÿ Map shows –Zip code locations colored by wind-induced residential building damage –Constrained by value, population and wind damage above thresholds

Summary Ÿ Deep Thunder is an integrated system that is –Usable forecasts are available automatically, in a timely, regular fashion –Illustrates the viablity of cloud-scale weather modelling to provide more precise forecasts of severe weather –Can be customized for different business applications and processes for safety, economic benefit and efficiency Ÿ Continued research and development –Improving quality of forecasts as well as product delivery –Adaptation of other research efforts to support operational applications –Multiple model forecast domains as platforms for development and collaboration Ÿ Future work –Adaptation and evaluation to other geographic areas –Enhanced workstation and web-based visualization, model tracking/steering and interactivity for both decision support and analysis –Improved computational performance and throughput –Extensions to still other models and data products –Customized interfaces, products and packaging for other applications (e. g. , emergency planning, aviation, surface transportation, broadcast, insurance, agriculture, etc. )