December 19, 2013 -- WHITE RIVER JUNCTION, Vt. – Maponics, a leading geospatial data provider, announced the release of Context™ Weather Scores, its newest product in the Context™ suite of lifestyle and behavioral analytics.
To build Context Weather Scores, Maponics engineers applied geographically weighted regression to model weather and climate conditions at the neighborhood level throughout the United States.
Other weather and climate models geared toward consumers report conditions at the local airport or nearest weather station. The Maponics model provides accurate information about temperature, precipitation, cloud cover and more at weather stations, airports and all areas in between.
“As part of our focus on providing data about real-world conditions, we wanted to accurately reflect climate at the neighborhood level – not just at weather stations, which are often placed in outlying sections of a metro region,” said Darrin Clement, Maponics CEO. “We’re delighted with the accuracy of results we obtained using a geographically weighted regression model for this product.”
Geographically weighted regression (GWR), which is relatively new, has been applied to numerous industries, including mining, oil and gas, crime modeling (including Maponics Context ™ Crime Scores), demography, regional planning, emergency service planning and risk management, among others. However, Maponics’ use of GWR for climate modeling on this scale has never been done so precisely or comprehensively.
To model climate normals, Maponics engineers used GWR and other methods to analyze the relationships among spatial variables that affect temperature, precipitation, dew point, heat index, wind chill, wind speed and cloud cover. Climate “normals” are the normal high, low, precipitation and other recordable weather conditions for a location for a 30-year period, in this case from 1981 to 2010.
Maponics derived their predictors from a variety of sources, including the National Oceanic and Atmospheric Administration (NOAA), the US Department of Agriculture, NASA and other state and federal agencies. They also used digital elevation models from the US Geological Survey to factor in the shape of the land, which has the greatest effect of all.
Maponics based their product primarily on the 30-year normals collected from thousands of weather recording stations. They also used as much supporting data as possible from satellites and other sources to inform their interpolations of temperatures, precipitation, winds, clouds and other variables.
“I evaluated each of the interpolation methods and tools available to us and I found the estimates produced using GWR best fit the observations,” said Cory Martin, Maponics engineer. “GWR methods also worked exceptionally well in flat and mountainous areas alike.”
The Maponics model is similar to the PRISM climate mapping system created by the PRISM Climate Group. Both give geographical weight to weather stations, but the Maponics model includes snowfall, whereas PRISM does not. Further, Maponics’ calculations are far more precise than other climate models. For example, the climate maps from NOAA show snowfall depths in bands of 3-inch increments, whereas Maponics’ model calculates snowfall depths to a fraction of an inch.
The Maponics Weather Scores product has a variety of use cases in the real estate, social media and travel industries, among others.
“Our Weather Scores product answers home buyers' questions like, 'What is the difference between living in the mountains versus choosing a home in the lower elevations of a particular city, like Phoenix?' The methods we used capture that level of granularity very well,” said Michael Villarreal, Context Product Manager.
Maponics is the leading provider of premium-quality, geospatial data and analytics that underpin today's location-based applications and services. Maponics creates geographic boundaries for areas where people spend their time and money, such as neighborhood boundaries, shopping boundaries, ZIP Code boundaries and school boundaries. Customers include many of today's best-known web, social media, mobile and real estate brands, including over 70% of the top real estate websites. 95% of all consumers who use social media interact in some way with Maponics data. The company is headquartered in Vermont. To find out more, visit maponics.com or call 1-800-762-5158.
Maponics and/or other noted Maponics related products contained herein are registered trademarks or trademarks of Maponics, LLC. All other registered and unregistered trademarks herein are the sole property of their respective owners.