Very hot summer's day in the Iberian peninsula

12 July 2017 05:00–13 July 00:15 UTC

Region

Spain

Satellites

Meteosat-10

Instruments

SEVIRI

Channels/Products

Infrared Channel

After a hot period in mid-June 2017, the Iberian Peninsula suffered a second heat wave in the second week of July, peaking on 13 July with a record maximum air temperature of 47.3 °C at Montoro, near Cordoba.

Last Updated

05 May 2023

Published on

12 July 2017

By Jose Prieto (EUMETSAT)

The definition of a heat wave from AEMET (the Spanish Meteorological Agency) refers to percentage thresholds exceeded in a number of stations, sustained for longer than three days and based on climatological values in July and August, the hottest months. In the last few years heat waves outside of those months seem to happen more frequently.

Enhanced Meteosat-10 infrared channel 10.8 µm, 12 July 13:00 UTC — Figure 1: Enhanced Meteosat-10 infrared channel 10.8µm, 12 July 13:00 UTC

Satellites do not sense air temperature close to the ground. There is a difference of a few degrees celsius between the air temperatures and the soil temperatures. Soil temperatures are easier to determine with the help of satellites, which provide continuous coverage of large surfaces. This soil/air difference varies with wind, Sun radiation and soil type, plus moisture or vegetation.

Upper panels: 3.9 µm-10.8µm differences for three Iberian locations W, C and G (Guadalquivir) on 12 July (left) and actual difference map near midday (right). Lower panels: Same, now for absolute channel 10.8 µm values. Mountain ranges and water bodies are easy to recognise in darker grey in the soil temperature map (right). — Figure 2: Upper panels: 3.9 µm-10.8µm differences for three Iberian locations W, C and G (Guadalquivir) on 12 July (left) and actual difference map near midday (right). Lower panels: Same, now for absolute channel 10.8µm values. Mountain ranges and water bodies are easy to recognise in darker grey in the soil temperature map (right).

For 12 July, the Meteosat-10 infrared animation shows the evolution of soil temperatures, as conservatively estimated from the brightness temperatures at channel 10.8µm.

Some central and southern peninsular areas absorb the solar radiation more efficiently than other regions, and reach soil values of up to 54°C for the pixel average, and obviously higher at many smaller areas.

The daily evolution of the ground temperature peaks around 13:00 UTC, as estimated from both the Meteosat-10 10.8µm channel and from the 3.9µm channel. The difference between 3.9µm–10.8µm peaks, on average, at 11:15 UTC (midday locally), as a result of the soil reflected component being at its highest around noon (see Figure 2).

For the individual channels, the peak happens about two hours later, at 13:00 UTC, as a consequence of soil thermal inertia. Furthermore, the peak in air temperature happens another two hours later, because of the poor conductivity of the dry air.

The fact that the mentioned difference between channels 3.9µm and 10.8µm is usually negative can be explained by the low emissivity at channel 3.9µm for dry areas, while the Sun's contribution is weaker than the hot soil contribution.

Figure 3: Heating and cooling patterns seen on thermal imagery of Meteosat-10 at 08:30 (left) and 18:00 UTC. Dry low areas take typically longer to cool than mountain or wet areas. The cooling inertia is a function of soil diffusivity, related to moisture and emissivity.

An illustration of the asymmetry between the morning warming and the evening cooling is provided on Figure 3, showing two moments in the day, morning and afternoon, with similar brightness temperatures on Meteosat-10.8µm. For different kinds of terrain, the cooling happens at different speed. In particular, the crops south of Guadalquivir (letter G in Figure 2) take longer to loose heat than hilly areas (letters W and C).

Figure 4 shows two proxies for soil classification. At first glance, both sets of images show a low correlation, indicating that the emissivity is not the only explanation for the variability (and maximum temperature forecasting). Air humidity could also contribute to the variation.