Reuters Climate Monitor

Loading today's forecast...

Difference from the 1961-1990 historic norm Based on UTC calendar days. Forecasts using local times will vary. See the box at the bottom of this page for more information.

≤-7°-5°-3°-1°+1°+3°+5°≥+7°

… Loading today's global summary…

… Loading global population exposure…

… Loading chart data…

… Loading monthly high-temperature history…

… Loading today's continent summary…

… Loading ranking data…

How this year's highs compare<br>… Loading chart data…

More from the Reuters Climate Monitor

Tracking US heatwaves

Tracking Western Europe's heatwave

Tracking Asia's heatwaves

About this data

The Reuters Climate Monitor shows where temperatures are unusually hot or cold in real time by comparing today’s conditions with what was typical in the past.<br>We start by establishing what normal used to look like. Using hourly temperature records from the ERA5 reanalysis dataset published by the European Centre for Medium-Range Weather Forecasts’ Copernicus program, we calculate the typical temperatures for each of roughly one million grid squares covering the Earth’s surface for each day of the calendar year.<br>Climate scientists typically call this kind of historic baseline average the normal temperature. Our historic norm is drawn from a 30-year window spanning 1961 to 1990, a standard reference period used by climate scientists. To reduce noise, each day’s reference temperatures are calculated using a 31-day rolling window — 15 days on either side of that calendar date — across all 30 years of data.<br>ERA5 is typically about five days behind real time. To compare current conditions, we use the latest run of the ECMWF medium-range control forecast, formerly known as HRES, which covers the same global grid.<br>Because the control forecast and ERA5 are generated by different models, they can produce slightly different readings for the same conditions. To address the issue, we apply a statistical correction — calculated by comparing the two models over a multi-year overlap — to put them on equal footing.<br>With the datasets aligned, we subtract the historic norm from today’s forecast for each grid square. The result is the anomaly reported by our map, which shows how much hotter or colder today’s high temperature is compared to what used to be the typical high on that date. Forecast highs are calculated over UTC calendar days, which may span parts of two local dates depending on location.<br>To calculate numbers for continents, countries and other regions, we collect every grid square that touches each area and compute an average. Country boundaries come from Natural Earth. Island territories far from a country’s mainland, such as the Azores or La Reunion, are counted separately.<br>Monthly figures are the average of each day’s regional high temperature over a calendar month. For the monthly historic norm, we average the same calendar month’s figures from 1961 through 1990. Until at least 15 days of HRES data are available for the current month, the charts continue to show the latest complete month. They then show a provisional average using available ERA5 data and the adjusted HRES forecast. During the month, that figure is compared with complete months from earlier years and may change as more days are added. It is finalized when complete ERA5 data become available.<br>For the United States, we also report the nine climate regions defined by the National Centers for Environmental Information.<br>For Europe, we report the NUTS-1 statistical regions developed by the European Union prior to the United Kingdom leaving the organization. Regions too small to be covered by more than one ERA5 grid cell are omitted.<br>For Asia, we report the regions defined by the group of scientists that produced the Intergovernmental Panel on Climate Change’s Sixth Assessment Report.<br>To estimate how many people are experiencing temperature anomalies, we match every square in the gridded climate data with a population count from LandScan Global, a 2024 dataset produced by Oak Ridge National Laboratory.<br>The population charts use isotype silhouettes from Wee People, an open-source typeface created by ProPublica and Alberto Cairo.<br>Our method was developed in consultation with more than two dozen climate experts, who provided guidance and advice. Special thanks are due to Alistair Hobday of CSIRO, Lukas Brunner at the University of Hamburg, Claudio Piani at the American University of Paris, Ed Hawkins at the University of Reading, Megan McCrary at Oak Ridge National Laboratory, Rebecca Emerton at the European Centre for Medium-Range Weather Forecasts, Julien Nicolas at Institut Galien Paris-Saclay, Harry Kennard at the University of Texas and Jonathan Chambers at Planeto.<br>We want your feedback on this feature and how it could improve. Contribute to our user survey and send any questions or comments to Reuters editor Ben Welsh.

Research and development

Ben Welsh and Casey Miller

Editing

Ben Welsh, Maurice...