MetStorm Rainfall Now Available as an Image Service via ArcGIS Server
WDT, which was recently acquired by DTN, has had a long standing partnership with MetStat, a hydrometeorological firm based out of Fort Collins, CO. Collectively, DTN and MetStat use the best data and technology to produce the best estimates of rainfall for the Continental US, and now that data is available as a set of image service from our ArcGIS Server so you can see rainfall accumulations for the last 1 hour up to the last 7 days.
Note — We’ll contour them soon for those that prefer feature service.
About the MetStorm Service:
MetStorm provides analysis of rainfall accumulations (often referred to as QPE — Quantitative Precipitation Estimate in the weather world) for running 1, 3, 6, & 12 hour time periods in addition to 1, 2, 3, 5, & 7 day running time periods. The MetStorm data are broken up into 2 time-enabled data services: one providing accumulations for time periods < 1 day and the other providing accumulations for the time periods that are > 1 day. The MetStorm product has a 1km resolution and updates hourly. Using REST endpoint to access the MetStorm image service, user can easily:
- Add the MetStorm data to web maps/applications
- Export images as GeoTIFFs or other formats
- Extract rainfall values at point locations within the gridded domain
What is MetStorm?
Without getting into gross detail, MetStorm is a proprietary process that we use to combine multiple data sets including rain gauges, radar, satellite, and terrain to get the best estimate of rainfall (QPE) across the Lower 48.
Why is MetStorm the best rainfall estimate?
Rainfall is a tricky thing to measure, yet wildly important to get it right. When used alone, all rain measuring techniques have their flaws:
- Rain gauges — Whether manual or automated, rain gauges only provide the measured rainfall at a point location. It’s fair to assume that locations very near to a rain gauge may have experienced the similar amounts of rain, but as you get farther from the rain gauge, the rainfall amounts will be less similar (see Tobler’s First Law Geography).
- Radar — Imagine using an instrument that’s 30 ft wide to measure something that’s a fraction of an inch in diameter 100 miles away. As you can imagine, there’s plenty of room for error. While what we can do with radar is nothing shy of amazing, remote sensors always have bias.
- Satellite — See above, but 100 ft wide instrument, over 22,000 miles away.
The best approach to calculating QPE is to combine all of these measurement techniques while also accounting for climatology and terrain. Simply put, MetStorm uses proprietary, patented dual polarized radar technology to quality control radar data. Dual-pol ensures we can correctly estimate the size/shape of a hydrometeor (in this case, a rain drop) because we can measure how wide and how tall a hydrometeor is (This is important. Here’s why). From there we can infer how much rain may have fallen across the US, but this is just an estimate based on remote sensing.
MetStat provides DTN with all of the rain gauge data they can get their hands on. Everything from top-of-the-line rain gauges at airports and other locations, to rain gauges that the average guy has in his backyard which require manual readings. It’s never safe to assume that all measurements are correct, so MetStat uses proprietary algorithms to flag and remove rainfall measurements where quality is suspect. Once possibly errant rain gauges are removed, the rain gauge values are compared to the radar estimates. The rain gauges are considered ground truth and are used to adjust the radar estimates. Once that bias is removed and the rest of the climatology and terrain secret sauce is applied, you get the best possible estimates of rainfall all across the US for every square kilometer regardless of whether there’s a rain gauge or not.
Have a need for the best rainfall estimates available?
- Agriculture
- Flood Control
- Oil & Gas
I’d e happy to set you up with a trial. Fill out this form, and note that you’re looking for the MetStorm GIS service.
