This is a repost from my post written in 2010 on Whatever-Weather.com.
In Meteorology we use models all the time. We use the NAM, GFS/GFSx, RUC and others on a daily basis. We know they are gridded models with multiple levels (e.g. 60 vertical levels for the NAM) and a larger grid resolution (e.g. 12km for the NAM). We compute multiple equations at each level for each grid point to use the output as part of our daily analyses. What comes out of each model on most days is a decent conceptual model of the atmosphere.
The atmosphere is only one part of a conceptualized real world. The world is made up of other features which can be related to the atmospheric models and other real time data. This is where Geographic Information Systems (GIS) come in. GIS relates real world features in either vector (e.g. points, lines, polygons) or grid form with a geospatial database on the back end. These features can be combined with weather data converted into the same form to provide real time and archived analyses. The sky is the limit for how diverse studies or real time applications you want to do.
GIS is in the middle of a transition from the mostly static desktop to dynamic server realm. It started out as a faster way to produce paper maps and is transitioning to server applications serve out real time relevant data to people who request it. The power is with the power of geoprocessing and map algebra in GIS. Geoprocessing allows you to run scripts manually or automatically to manipulate different data sets together through map algebra, and in a specific order to get your final analysis. This can then be sent out to clients to help them in their decision making processes.
Let’s consider a flooding example. With spring coming, rivers will be rising. Every River Forecast Center across the country will be making forecasts about expected river crests. The expected forecast crests are then sent to Emergency Managers who can then use GIS to get a rough estimate about damages from the flooding based on house location and height of the ground (e.g. using Digital Elevation Models, DEMs).and how to use their resources for flood protection for the general public. A real world example of this is the Flood Forecast Display for the Fargo, North Dakota/Moorhead, Minnesota region along the Red River of the North.
The National Weather Service (NWS) offers both static and live dynamic GIS data, much of it for no cost. You can convert most static data or data streams which are not shown at the NWS site into useful GIS formats for analysis with non-meteorology data. The NWS also offers a converter for National Digital Forecast Database (NDFD) to convert GRIB2 data into useable GIS formats, either through a standalone application (tkdegrib.exe), or with automatic scripting (degrib.exe).
Current solutions from ESRI with their ArcGIS Desktop and ArcGIS Server applications are the easiest to configure and develop applications and solutions. There are also open source versions of GIS, like Mapserver which is similar to ArcGIS Server to serve data. They work well but often have a higher learning and configuration curve. The forecast suite F5Data also utilizes a GIS engine to show current and model variables.
Geospatial technology through GIS is the technology of the future in relating the effects of weather to the public and private industry. It will continue to progress to almost real time relation of the weather to the world so people can remain safe with ever changing conditions.
Flood Forecast Display Tool:
National Weather Service GIS Data Portal:
National Weather Service NDFD GRIB2 Decoder:
©2010-2012 Charles Schoeneberger