| title | grand_parent | parent | layout | nav_exclude | scripts | examples | |||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Land Surface Temperature Backward Average |
Planetary Variables |
Land Surface Temperature |
script |
false |
|
|
The Land Surface Temperature Backward Average is a method to reduce data gaps and measurement noise in the Land Surface Temperature (LST) data. Depending on the requirements, we can choose a lookback period, for example 20 days. The 20-day backward average of LST for day n is the average of LST over the 20 days preceding day n. We compute the backward average using all available measurements within this 20-day period, and therefore, we do have a valid value for every day, except in case of prolonged data unavailability, such as during long frost and snow periods.
The Land Surface Temperature Backward Average is suitable for applications where long-term temperatures are more relevant than daily fluctuations. The moving average operation reduces day-to-day variations and in the resulting time series, seasonal and longer-term changes can be easily detected. It can be used for monitoring drought risk, yield forecasting and analysis of climate change.