“The climate impacts the landscape, but the landscape also shapes the climate: exchanges at the surface of the soil and plants influence what happens in the atmosphere.”

Fanny Léglise. What are the characteristics of the climate and territory of Gironde?

Yves Brunet. The Gironde has an altered oceanic climate. “Oceanic” because winters are mild, with few days of frost and temperatures rarely dropping below freezing. Summers are warm without being scorching. The prevailing winds from the southwest to northwest bring ocean air. Rainfall is relatively abundant¹. “Altered” for two reasons: the department can experience relatively high temperatures with a high number of hours of sunshine² for a non-Mediterranean region. The Atlantic coastline is linear and perpendicular to the prevailing winds, with a west-east gradient of temperature, humidity, and precipitation. The coastal strip is sunny but cooler and more humid. There is a gradual “continentalization” towards the east. It is a region particularly exposed to the risk of fire, with the presence of large forests and frequent summer droughts.

The Gironde region is relatively unaffected by relief and has a dense hydrographic network. It is highly exposed to rising temperatures due to climate change. While it remains difficult to establish regionalized forecasts for climate change, a few major trends are emerging, which can be found on AclimaTerra³. Two reports – in French – were published in 2013⁴ and 2018⁵ concerning the future of Aquitaine.

FL. Among the contributions to Nouvelles saisons, architects from cdlt exhibited a rain gauge alongside an editorial dedicated to clouds in Aquitaine. They encouraged us to reflect on what connects us all: the atmosphere. As a micrometeorologist, what does it mean to you that climate can be the subject of an architecture exhibition?

YB. Many elements are related to water flow, shade, vegetation, and thermal comfort in architecture, urban planning, and landscape design. I imagine that we don't build the same things here as we do in northern France or in the mountains... I think it makes perfect sense to take these factors into account. It seems to me that this convergence of disciplines is a recent development; it wasn't previously expressed in such a meaningful way. The scale chosen in Nouvelles saisons, that of the Gironde region, is an “intermediate” scale in terms of climate. In meteorology, we talk about the “sub-regional” or “sub-meso” scale (a few dozen kilometers). At an even finer scale, we also talk about the “landscape scale”: this is the scale that is perceptible to the eye, that of land use, agriculture, and land management. It therefore makes a lot of sense.

FL. What is your research focused on and how does it relate to the interests of those involved in urban development?

YB. I am a micrometeorologist. I study what determines the weather over short periods of time, locally or regionally. This sets me apart from climatologists, who work on long series of years, and even more so from paleoclimatologists, who study tree rings or fossilized pollen to understand even longer time scales. To understand the difference between climate and meteorology, take a look at your wardrobe: the clothes in it reflect the local climate, and what you wear today is determined by the weather.

Prior to the urban issue, we worked with colleagues from Météo France to study the impact of land use – particularly the presence of forests – on cloud cover⁶. A Dutch study based on ten years of satellite data on the Landes forest⁷ showed that in summer conditions, significant cloud cover (small fair-weather cumulus clouds) often occurs above the forest – following the shape of the Landes massif – with the exception of the coastal strip. This study covers the five years preceding Storm Klaus in 2009 and the five years following it: there was a very clear decrease in cloud cover after approximately half of the trees in the heart of the forest fell. Thus, a storm is not “just” a matter of falling trees; it has a longer-term impact on the local climate.

We were able to reproduce these phenomena using numerical simulations with current modeling techniques⁸. We then set up simulations based on assumptions about land use: what would happen if we replaced the Landes forest with 0% forest, or 20%, etc., up to 100%? The results show that with less than 30-40% forest cover, there is little impact on cloud cover; above 40%, each time trees are “added,” there is a net effect on both the appearance of clouds and their thickness; above 70–80%, there is hardly any effect. In other words, in a forest such as that of the Landes, with intermediate density (roughly half trees and half crops), the local climate's reaction to deforestation or afforestation is very sensitive.

FL. Does this mean that a virtuous cycle is taking place: clouds would remain stationary above wooded areas, which would benefit from increased rainfall and thus grow better?

YB. There is a self-limiting factor to this phenomenon, linked to the absorption of radiation by the forest. As the forest has a lower albedo⁹ than crops, it absorbs solar radiation more effectively and this surplus helps to generate upward currents known as “convective currents,” which transport moisture to an altitude where it can condense. But when clouds accumulate, they reduce the amount of radiation reaching the ground: the convection engine is weakened, which regulates the cloud formation mechanism.

Furthermore, we have only considered fair weather clouds so far. To take this further, we would need to look at precipitation and anticipate the formation of rain in different landscape contexts. All of this is still only a theoretical model at this stage. However, more and more researchers are working on scales smaller than those of “classical” meteorology: some are focusing on urban microclimates, while others are analyzing, for example, on scales of tens or hundreds of meters, the microclimates generated by hedges bordering agricultural plots, or by the topography of the landscape (valleys, hills, etc.).

FL. What does your research tell us about the relationship between land cover and the atmosphere?

YB. Some elements of the landscape influence the climate in a very localized way. In the early morning, under certain conditions, you can guess the presence of rivers, particularly the Garonne and its tributaries, by the bands of fog. At INRAE, we have been working on the specific case of the Ciron Valley, which joins the Garonne downstream from Langon. This valley has ecological characteristics that are unique for our latitude and altitude, with the long-standing presence of a beech forest: under “normal” conditions, this tree species grows either further north or further south at higher altitudes. Its existence in the Ciron Valley can be explained by the presence of “gorges” – a term that should be taken relative, as they are only about ten meters deep. The 50 or so temperature and humidity sensors we have installed there have shown that this difference in altitude is sufficient to create a 3-4 degree temperature difference between the bottom of the valley and the plateau. It is less cold in winter and less hot in summer¹⁰. This natural “buffering” effect of the forest is exacerbated by the topography, which creates local conditions similar to those usually found at an altitude of 700 meters. This beech forest is evidence of the existence of a very local “climatic refuge.”

More generally, we are trying to define the scale at which land use affects the “boundary layer,” i.e., the first two kilometers of the atmosphere. A minimum size of forest, or heterogeneity in the landscape, is required to have an impact on cloud cover. The existence of a large photovoltaic farm project (measuring approximately 3x3 kilometers in its original version) near Saucats has raised questions among neighboring winegrowers about the possibility of effects on the microclimate, particularly cloud cover. There is currently no “ready-made” answer to this question. All we can say is that an area less than 10 kilometers wide, in the direction of the wind, is unlikely to affect cloud formation.

FL. Does the nature of the soil in our cities influence cloud cover and, as a result, the climate?

YB. Clouds form thanks to the water on the surface and a mechanism that allows water vapor to rise (thermal convection, wind, turbulence). In cities, there's often a lack of water on the surface, as the ground is mostly impermeable and water is drained away through various systems. Vegetation plays a small role in supplying water to the atmosphere through leaf transpiration, but the effect is quite limited: cities often experience a “heat island” effect due to the high absorption of radiation by urban materials and their inertia, and the low proportion of this radiation used for evapotranspiration. This heat island is primarily a nocturnal phenomenon. In the 1970s, when the first infrared satellites capable of measuring surface temperature appeared, cities stood out on nighttime images. They absorb heat during the day and release it at night.

In Nantes, the École Nationale Supérieure d'Architecture and the École Centrale¹¹, as well as the Centre National de Recherches Météorologiques in Toulouse, are interested in urban microclimatology, particularly in relation to building thermal performance and street microclimates. Researchers are applying fluid mechanics models to study the effects of trees in a street on the microclimate and on pollution from cars. We know that vegetation captures pollutants, but the effect on air quality depends on the layout of the trees. In a “street canyon,” i.e., a straight street lined with tall buildings, once a certain level of tree coverage is reached, exhaust gases can become trapped under the vegetation, leading to increased pollution for pedestrians.

At INRAE, we are also interested in urban microclimates at all scales: that of pedestrians with tree shade, that of streets with rows of trees, that of blocks, neighborhoods, cities, and metropolitan areas. The larger the scale analyzed, the greater the effects on the atmosphere. There is a real link between the two. Other studies are attempting to quantify the link between the percentage of green spaces and street microclimates.

FL. Could an increase in the number of cooling islands have a multiplier effect on a larger scale?

YB. On a city scale, and for a given fraction of the area occupied by vegetation, we might ask ourselves what impact its spatial distribution has. For example, will the cooling effect be greater with lots of small gardens, lots of small parks, or just a few large parks? This point is the subject of active research and we do not yet have a clear conclusion. We only know that an urban island of a given surface area has a certain radius of action. A vegetated area with a diameter L affects the microclimate along an outer distance of the order of L, with a fairly rapid variation in temperature and humidity, which decreases exponentially. A park 100 meters long will no longer have a significant effect on the microclimate 100 or 200 meters away.

By studying the impact of trees on the climate, we can better consider land use planning and management to promote, for example, a reduction in the heat island effect. The type of studies mentioned above has led some scientists to propose planting large strips of trees near cities, on the side of the prevailing winds, which could impact the urban microclimate by limiting the rise in summer temperatures, or even by creating shade over cities, or by giving rise to precipitation further downstream. The calculation tools now available make it possible to simulate such effects. The studies to be conducted could thus lead to spatial planning perspectives: how and where to reforest, for example? These are very interesting prospects, which will require interdisciplinary collaboration between atmospheric scientists, tree physiology specialists, foresters, planners, landscape architects, local authorities, and others. Everything needs to be organized!

Interview by Fanny Léglise on Novembre 3rd, 2025.

1. Between 800 and 900 mm per year in Bordeaux, or approximately one day in three with rain. ↗
2. Over 2,000 hours per year in Bordeaux, 2,200 in Arcachon. ↗
3. www.acclimaterra.fr, Regional Scientific Committee on Climate Change. ↗
4. www.acclimaterra.fr/wp-content/uploads/2015/10/LES-IMPACTS-DU-CHANGEMENT-CLIMATIQUE-EN-AQUITAINE-110Mo.pdf ↗
5. www.acclimaterra.fr/wp-content/uploads/2018/05/Rapport-AcclimaTerra.pdf ↗
6. Gradual cloud cover spreading across the sky. ↗
7. Teuling A.J. et al., 2017. Observational evidence for cloud cover enhancement over western European forests. Nature Communications, 8:1-7. https://doi.org/10.1038/ncomms14065. The Landes forest is an interesting example because there is no topography to interfere with the recorded signals. ↗
8. Noual G., Brunet Y., Le Moigne P., Lac Ch., 2023. Simulating the effects of regional forest cover changes on mid-latitude boundary-layer clouds. J Geophys Res. Atmos., 128:e2023JD038477 (http://dx.doi.org/10.1029/2023JD038477). ↗
9. Albedo is the reflectivity of a surface, i.e., the ratio of reflected light energy flux to incident light energy flux. ↗
10. Ogée J., Walbott M., Barbeta A., Corcket E., Brunet Y., 2024. Decametric-scale buffering of climate extremes in forest understory within a riparian microrefugia: the key role of microtopography. International Journal of Biometeorology, doi.org/10.1007/s00484-024-02702-9. ↗
11. Musy M. (coord.), 2014. Une ville verte. Les rôles du végétal en ville. Éditions Quae, in French. ↗