Home

[April 9, 2020: This wiki is under a lot of development right now]

Gridpp a is a tool for performing statistical post-processing of gridded weather forecasts. It consists of a library of commonly-used post-processing methods and a command-line tool that applies these methods to forecast fields in NetCDF files.

Gridpp is written in C++ but offers python bindings to the functions in the library. The tool is used at MET Norway to produce operational weather forecasts for Yr (https://www.yr.no).

Features

• Methods for downscaling a forecast from a coarse grid to a fine grid, such as bilinear or nearest neighbour methods.
• Methods for calibrating a downscaled grid, such as quantile mapping.
• Computationally efficient neighbourhood methods to compute neighbourhood min, mean, max, and any quantile.
• Functions for diagnosing variables, such as relative humidity from temperature and dewpoint temperature.
• Data assimilation methods such as optimal interpolation (OI) to merge observations and gridded forecasts (deterministic or ensemble)
• Efficient data structures for nearest location lookup in a vector or grid of locations

This wiki includes documentation and examples of how to use the gridpp command-line tool and the python interface to the gridpp library.

News

Date	News
June 15, 2020	A prototype version of gridpp library v0.4.2 is available.
April 9, 2020	Gridpp is being restructured to offer an API to its core functions.