Posted on Wed, 15 Mar 2017 by John
We have made improvements in how we ingest historical temperature data, and other sensitive weather data. We have also improved our estimations for missing historical temperature data, precipitation data, and other weather data parameters!
Our historical weather API now includes dry adiabatic adjustments when backfilling temperature data. This adjustment allows the backfilled temperature data to account for elevation differences between two stations by using the dry adiabatic lapse rate. In other words, for every 1000 meters of elevation difference between two stations, the temperature is adjusted by (+/-) 9.8 degrees Centigrade.
This historical temperature data adjustment method allows for us to derive a much more accurate estimate of what the historical temperature actually was at a given location with missing data!
Our historical weather API now includes more accurate weather condition data for stations that only report 6 hourly precipitation. By using a combination of cloud coverage, humidity, and surrounding stations conditions we are able to approximate more accurately the weather conditions in the absence of hourly precipitation data!
Six hour precipitation data is now backfilled correctly along with hourly precipitation data. This is a critical feature for those that are looking for historical precipitation data, as a majority of weather stations only report precipitation at 6 hourly, or irregular intervals.
Our historical weather API now includes more accurate handling of missing pressure observations. Backfilled historical pressure data is adjusted for altitude differences between stations using the barometric formula. This allows for a more accurate estimation of missing pressure data.