Interpolating Land and Sea Points
Interpolation of 2m temperature or 10m wind over or near a coastline currently uses the land-sea mask to decide the whether the four surrounding model grid points are land or sea points. This determines whether the interpolated value should be regarded as a land or sea point. In this way there will be no undesired smoothing of gradients along coastlines. There can be difficulties with strong gradients along coasts, near small islands, or in mountainous regions. There were historical reasons, related to resolution, for using the land-sea mask in interpolation procedures. There is much less need for this now that the IFS model resolution is much higher. The interpolation package MIR does not use a land-sea mask when interpolating data to another resolution. This is because the data will already have taken the land-sea mask into account during processing at the appropriate resolution of the IFS. However, interpretation of the IFS land-sea mask by MIR shows a few points change in nature from land to sea or sea to land. This very slightly modifies the display of geographical features.
At grid points along coastlines the marine influence may be overestimated. Statistical interpretation schemes (e.g. Model Output Statistics (MOS)) can be beneficial, particularly for temperature forecasts. Avoiding mis-assignment of radiation at coastal land sites is also important.
Additional Sources of Information
(Note: In older material there may be references to issues that have subsequently been addressed)
- Read "Reducing surface temperature errors at coastlines" for more details.
Overview
Community Forums
Content Tools