Interpolation of weather variables for point locations. #549
Replies: 2 comments 3 replies
-
Hi, thanks for asking. I downloaded era5-land for the entire grid (global). Afterwards, I converted it into a time-series for each location (transpose). The API picks the nearest grid-cell. However, if the grid-cell is has a different surface elevation than the desired coordinate, it will analyse the surrounding grid-cells and select a more suitable elevation. This drastically improves accuracy in complex terrain. Similar behaviour applies to coastal areas. If you do not want to have this dynamic selection, use I do not do any spatial interpolation between grid-cells. Interpolation can reduce accuracy and introduce unwanted artefacts. |
Beta Was this translation helpful? Give feedback.
-
Interesting to know. I though it was a good practice to do spatial interpolation between grid-cells. Could you elaborate on the unwanted artefacts and accuracy reduction? |
Beta Was this translation helpful? Give feedback.
-
Hi @patrick-zippenfenig ,
I have a question behind the logic of point data. Let's take the historical api as an example. And specifically the era5-land-hourly data. When downloading this data using the CDS API, you need to provide a rectangle for which you want climate data and you will receive data on a grid with a resolution of 0.1° x 0.1°. How did you go about transforming the gridded data from CDS to point location data, since using the historical API of open-meteo you can only specify a point location? Did you use some interpolation techniques or simply used the nearest neighbor? Also, at what resolution did you download the data from era4-land-hourly?
Kind regards,
Jens
Beta Was this translation helpful? Give feedback.
All reactions