Skip to content
Rasmus E. Benestad edited this page Sep 13, 2023 · 14 revisions

Welcome

Welcome to the OpenClimateDataPrototype wiki! These pages will hopefully explain what the prototype Prototypeapplication is about and help you to install it on your local computer. One example of the app can be found at this link. There is also a screencast (video) that gives a brief tour/demonstration of how to use the app on YouTube.

The app can be used in two ways: (1) as a web portal for presenting station data or (2) as an app on a laptop/server for examining daily station data stored in a netCDF file (see demo below and the screencast above).

Principle

This prototype lets you explore climate data from station records. It starts with an overview of the data and lets you then zoom into location of interest and presents you with a time series. It is designed to be flexible so that it can be applied to anywhere in the world - based on whatever data file is provided. It tries to be smart so that it extracts information required for the operation from the netCDF files. It is designed to be fast by only extracting the data it needs at any time (summary statistics, metadata or a chosen time series) from a local netCDF-file (making use of the direct access functionality). The use of local netCDF files also seperates the app from the database, which may be beneficial security-wise. The prototype can run on a local computer or as a web-based service.

Keywords for this prototype are: user-friendliness, flexibility, smart, fast, and secure. However, I don't know how well it scales if there is a massive user demand (needs to be tested).

Below is a schematic that shows the structure of the prototype, with three R-scripts which are the main part of it and how they relate to the data. There are also some shell scripts for assisting the set-up and management of the prototype. The data comprise netCDF files which are supposed to be generated by extracting station data from a database.
Schematic of the prototype

Description

The prototype is mainly a set of computer code that runs as an R shiny app and extracts information from climate station data stored as netCDF files. The app comes with a set of demonstration data from Met Norway, and is also set up to work with station data from European Climate and assessment Data (ECA&D) and the Global Historical Climate Network (GHCN).

Metadata is part of the analysis, and the prototype makes use of the metadata stored in the netCDF file (header as well as as variables) in the smart approach where the files and data are interpreted and visualised.

The engine behind the app is the R-package called esd which is available from github.com/metno/esd.

The prototype is made up of a code part (set-up: global.R; user interface: ui.R; and processing: server.R), a data part, and scripts (*.sh) to update the data and do some house keeping.

Some extra work needs to be done to make this app working with various databases. All that is needed is to read the data from the database into the R environment and use esd to save these in the netCDF format (a hybrid CF convention with some extra summary statistics) in the local data folder. This can be set up as a crontab job so that the data is up-to-date on a daily basis.

Security

See Mastering shiny.