Frequently Asked Questions

Frequently Questioned Answers

Sensor Point^tm Model Future Weather Plus^tm Forecast Data

Why would anyone use Future Weather Plus forecast data rather than free government forecasts?

Support for the National Weather Service’s (NWS) hour-by-hour forecast data is limited. Currently, it is available in experimental form at several sites in the US; however, there is no central distribution site. No international support is available. No customer service is provided. The NWS product is experimental and funding to extend it to other regions of the US has been inconsistent and remains uncertain.

What about other free online forecasts?

“Free” data comes with no customer service or technical support. It is also restricted to non-commercial use. No verification or reliability statistics are shown; free data is provided “as is”.

How does Future Weather Plus forecast data differ from that offered by other commercial providers?

Competitor forecast data is updated once each six hours. Sensor Point Future Weather Plus forecast data, in the form of an hour-by-hour time series, is updated once per hour. Forecasts are generally available for any site accessible from the Internet, including private stations.

How does Future Weather Plus forecast data performance compare with competitor forecast data?

Competitor model data tends to have a rather large “latency” compared to Sensor Point Future Weather Plus forecast data. This is partly because updates are provided only once each six hours. Since fresher data is usually better, Sensor Point model forecasts perform dramatically better, especially within the first 12 hours. For applications that require high precision and accuracy within this window, this should make a big difference. Contact us for details.

How does Sensor Point model make this happen?

Government models are very accurate, but are area – rather than location – oriented. The Sensor Point model’s forecast process uses station observations to “localize” government forecast model data. Contact us for further details.

Sensor Point^tm Model Past Weather Plus^tm Data

Why would anyone use Past Weather Plus historical data rather than “official” government raw station observations?

Raw station observation data present problems for some users; for example, sensors are relocated, changed, land use patterns for areas surrounding the station can change. These factors can cause station data to lose “statistical stationarity” and other biases. Raw surface observations also can be missing and contain obvious entry errors. This is true for both archived government and private-station data.

How is the Past Weather Plus historical data set different from “Cleaned” data offered by other providers?

Cleaned observations are data sets that have been made complete and debiased using various numerical methods. These methods may be statistical or a combination of statistics and model data. Past Weather Plus data uses station observations to localize government reanalysis model data. 

How does Sensor Point model make this happen?

The Sensor Point model creates hour-by-hour “re-forecasts” in the past as it assimilates any historical government observations that are available for that time. Raw historical observations, especially hour-by-hour observations, are often incomplete. Besides removing biases and preserving statistical stationarity, the Sensor Point’s Past Weather Plus process restores the information contained in missing historical observations.

What is Weather Analytics’ source of historical raw surface observations?

Quality-controlled surface station observations are directly provided by Weather Source, LLC. The US government archives raw observational data at its center in Asheville, NC. International agreements make raw station observations from foreign weather stations available. Once obvious entry errors are removed, they are made available to Weather Analytics’ Sensor Point model system.