Automatic Weather Stations (AWS) collect temperature, humidity and atmospheric pressure (and wind direction and velocity in some locations) from various locations throughout the world. Data from this system is used to record, and subsequently extrapolate, climate patterns. But is there a problem with the accuracy of the data collected by individual AWS installations because of localized urban conditions and features? Can smoke from a nearby BBQ distort readings? Can air conditioner exhaust distort them? Can heat from asphalt?
Image: a remote weather observation unit – click for larger image.
In what might be a major contamination of weather data, weather sensors can be influenced by the effects of surrounding human activity and artifacts. These factors will skew the data in some locations because of unnatural heating and distortion of natural climatic conditions. If you bias temperature readings in some locations, it would be impossible to rely on that data unless it has been normalized to data collected in undeveloped areas.
A Northern California blog, What’s Up With That?, identifies four cases where Urban Heat Island (UHI) effects are likely distorting temperature data.