Climate Diagrams for Any Location on Earth Experimental

The diagrams use the free global dataset CHELSA (Climatologies at high resolution for the earth's land surface areas). These high-resolution data cover the period 1981-2010, and pixel size is less than one square kilometer.

Karger, D.N., Conrad, O., Böhner, J., Kawohl, T., Kreft, H., Soria-Auza, R.W., Zimmermann, N.E., Linder, P., Kessler, M. (2017). Climatologies at high resolution for the Earth land surface areas. Scientific Data. 4 170122. https://doi.org/10.1038/sdata.2017.122

The optional current-year / previous-year overlay (dark red/blue and light grey bars and lines) uses a second, independent dataset: ERA5-Land, a reanalysis produced by the Copernicus Climate Change Service (C3S) at ECMWF. Unlike CHELSA's fixed 1981-2010 climate normal, ERA5-Land estimates actual monthly conditions for the current and previous year by combining a weather model with real observations. The most recent 2-3 months are provisional ("ERA5T") and may shift slightly once the final, quality-controlled version is published a few months later.

Muñoz Sabater, J. (2019): ERA5-Land monthly averaged data from 1950 to present. Copernicus Climate Change Service (C3S) Climate Data Store (CDS). https://doi.org/10.24381/cds.68d2bb30

ERA5-Land's native grid is about 9 km (0.1°) across — much coarser than CHELSA's ~1 km. Used as-is, that coarser grid would smooth away local terrain effects CHELSA captures (a valley town reading the same as the hills around it, for instance). To avoid that, every point is delta-downscaled rather than read directly off the ERA5-Land grid: CHELSA's fine sub-pixel pattern within that same ERA5-Land grid cell is used to adjust the current ERA5-Land value — the gap between the CHELSA normal at that exact point and the CHELSA average across the whole cell is added, for temperature, or applied as a ratio, for precipitation (to avoid negative values). This keeps CHELSA's local detail while still reflecting current, rather than long-term-average, conditions.

This overlay is an experimental addition and may change. Treat it as an approximate sense of how this year compares to the long-term normal, not a precise station-level measurement.

The diagram uses D3.js in an ObservableHQ notebook (Climate Diagrams). The map is a Leaflet map, geolocation is done by LocationIQ.

There are also other diagram variants — a Walter-Lieth climate diagram (the classic scientific diagram form: 1:2 temperature/precipitation scale, compressed above 100mm, with humid/arid shading, described on Wikipedia) and a Klimogramm (monthly temperature/precipitation pairs plotted as a chronologically connected loop).

The climate classification (Köppen-Geiger, Cannon, Trewartha, Defaut 1996, Thornthwaite-Feddema, Whittaker) is done using pyZonae by Didier M. Roche, Didier Paillard and G. Bonneroy, a modernized merge building on the original pyKoeppen.