City Data Package Description

It is a data package of 1 weather variable or 40 indicators with time series of monthly or daily values from 1951 to 2100, as simulated by 17 to 21 models (according to the variable) under 2 greenhouse gases emission scenarios (RCP4.5 and RCP8.5). Data are time series from a single site, corresponding to the center of a model grid box of 0.25°x 0.25° (about 25km x 25km).

They are "Ready to use" because they are suitable for direct use in climate change impact studies and physical risk assessments. They are obtained through statistical processing with advanced methods to remove model biases, increase spatial resolution and control output quality (more here on the data processing).

Recommended use
Ready to use data are useful for quick exploration of future climate conditions at a specific location. It is intended to inform about possible trends and changes to support qualitative assessments. For example for establishing an adaptation strategy, it provides the background information allowing you to infer a general understanding of the extent to which climate change may affect your region of interest (e.g., allows you to assess how temperature or precipitation may evolve allowing you evaluate societal impact).

Are the observation data from measurement at weather stations ?
Not exactly. What we call “observations” are actually “Reanalysis” data. Reanalysis is a technique that combines climate modeling and measurements from all kinds of weather instruments. In this dataset, the reanalysis data is called ERA5 and available from the Copernicus Climate Change Service provided through the Climate data store.

What are the limitations ?
The data is comparable but not exactly similar to what a weather station measures. If you compare the present observation data with your favourite historical weather station data they should be close but they will probably not match because they do not represent the same spatial scale (a square meter compared to several hundreds of square kilometers).

Do not expect the model date to be comparable on a day to day basis between models or with the provided observations. It is the statistics over a 20 to 30 year period that should be compared not specific days or months.

Do not expect the data to account for small scale processes that might alter the values given here. For instance, to account for the Urban Heat-Island effect further modeling at higher spatial resolution is required. For these, we recommend contacting specialized environmental consulting firms.

The data is available from our Data Shop for more than 4,300 cities worldwide that have more than 100,000 inhabitants.

Was this article helpful?
Thank you!