Wiiat Deforestation Does To Climate Within A Region

What will happen if a forest is removed and replaced with much more open vegetation, such as grassland or fields of crops? In a general way, there will be two competing effects on local climate. First, albedo will be greater over the more open grassland or cropland with patches of lighter soil between the leaves. This will tend to cool down the surface because solar energy is reflected straight back to space. However, the smoother surface of a grassland or crop cover—and the smaller total

Tropical Rainforest Layers DiagramTropical Region Solar Energy

Figure 6.3. (a) In the tropical rainforest, loss of latent heat uptake and roughness dominates and deforestation is predicted to result in a regional temperature increase, (b) In boreal forest (which has less evaporation) the albedo effect dominates upon forest clearance, producing cooling. Source: Author.

Figure 6.3. (a) In the tropical rainforest, loss of latent heat uptake and roughness dominates and deforestation is predicted to result in a regional temperature increase, (b) In boreal forest (which has less evaporation) the albedo effect dominates upon forest clearance, producing cooling. Source: Author.

that what the model says is valid. For example, it would be ideal to have good time scries of temperature and precipitation data from all across the eastern USA dating back to the time of the first European settlements in North America. Then, we could compare the "before" and "after" for deforestation in the region. However, no such data scries are available because no-one was collecting climate data back then, so we must look to validate the models in other ways. Fortunately, there arc observations from the present day that back up parts of the general picture from models about the effect of forest cover on climate in North America. This comes in the form of an unintentional sort of "experiment": What would happen if you have one region mostly covered in crops, and another region next to it mostly covered in forest? For mid-latitude regions, such as the USA, the models would say that compared with forest the cropland area will tend to be cooler during the peak of the day, bccausc the open cropland reflects much more sunlight back into spacc, preventing it from heating up so much under the sun. At night, temperatures in spring tend to cool to about the same level, because there is not much difference in the water vapor content of the air at this time of year, whatever the vegetation cover. In the tropics, the greater sun's heating combined with reduced transpiration from leaves would tend to have an opposite effect, making the cropland hotter, but in these mid-latitude areas the effect of albedo tends to be more important, Comparing observations of the daily temperature range in the Midwestern USA (where there is hardly any forest and mostly crops) with the eastern USA (where there is more forest than crops) shows that they differ in just the way that the models would prcdict. In the Midwest, the daily range during spring and autumn is less than in the forested east, because daytime temperatures in the Midwestern croplands do not rise as far from their night-time level. The east west difference is greatest in the late spring when the eastern forests have leafed out but the Midwestern crop plants are still small seedlings. At this time there is a lot of bare soil in the Midwestern fields reflecting sunlight, while the eastern forests are already dark and heating up. This agrees with what one would expect from the models: clear away a temperate forest and replace it by cropland, and things will be slightly cooler on average.

Another interesting set of observations that agrees with the expectations from models is the trend in this east west temperature difference over the last century, recorded from climate stations. Since the late 1800s, forest has been spreading back over abandoned croplands in the eastern USA. Yet, in the Midwest there is if anything less forest than there was at the turn of the 20th century, as agriculture there has intensified. The models would predict that the east-west contrast in spring temperatures would have increased as the east became more forested; and indeed that is what has happened.

The USA might also have undergone changes in rainfall patterns due to changes in forest cover. A modeling study by Copeland and colleagues suggests that the clearance of southeastern coastal plain forests in the USA. and replacement by-cropland since early colonial times, has shifted the peak area of rainfall southwards. Previously, the model suggests, the rainiest place in the region was the Appalachian Mountain belt. But now—both in reality and in the model—the most rainfall occurs over the northern edge of the coastal plain, at the boundary between cropland and forest. This hypothetical shift resulted from increased atmospheric upwelling over the sudden discontinuity in the landscape between forest of the mountains and rolling Piedmont country to the north, and cropland to the south. The fact that the maximum rainfall now occurs just where the model says it should is an encouraging result, which suggests that modelers arc getting things about right.

The effects of tropical deforestation on climate have occupied environmental scientists for several decades now, fueled by both old traditional concerns and the new results of climate modeling. Because deforestation in Amazonia has been happening so rapidly in recent decades, much of the scientific work in observing and modeling forest feedbacks on climate has been concentrated in this area. Various projects have compared the local climate, in areas that had recently been deforested for ranch lands, with adjaccnt areas of intact forest. These field observations showed that locally cleared areas tend to have an increased daily temperature range, and an increased daily range in humidity, with a peak of temperature and dip in humidity in the middle of the day. However, the overall average temperature and humidity did not change much. These were local-scale studies, but what would happen if Amazonia was deforested on a much broader scale, with all of its forest replaced by grasslands? Computer models that have simulated this scenario of widespread destruction suggest that there would be increased temperatures, less evaporation of water from vegetation and less rainfall.

The temperature increase in a deforested Amazon, around 1.4 C on average, would be due to less latent heat being taken up into transpiration from leaves, plus evaporation from rainwater sitting on the top of the leaves. The leaves of the grassland that might replace the rainforest cannot rival a whole forest canopy as an evaporative surface, and the turf of a grassland cannot match the forest's dense meshwork of roots and spongy soil organic matter for holding water. Note, however, that even though the temperature in Amazonia would go up, the world as a whole would get cooler as a result of this deforestation. This is because the latent heat that cools Amazonia is ultimately an important source of heat to the high latitudes (see the section below, on remote effects of deforestation).

Models predict that if all the Amazon forest was cleared, rainfall inland in the Amazon Basin would decrease by about 20%, enough to make things too dry for forest in some of the more climatically marginal areas of the rainforest. The main reason for this is the loss of recycling of water vapor within the rainforest region. When rain falls on intact forest, much of it is caught by the root mat and soil, and eventually taken up by the trees and evaporated from the leaves of the canopy. 'The moistened air then drifts further inland, where it can once again give rain that nourishes the forest. If a forest cover had not been in place, much of this rain could have run straight off into rivers and down to the sea, and not recycled. In the interior of the Amazon Basin, a large part (around 50%) of the rainfall depends on this recycled water vapor from forests elsewhere in the Basin. A grassland cover could recycle some of this rainfall too, but not nearly as effectively as forest.

So far, no decrease in rainfall has been detected in Amazonia, although only a relatively small proportion of the region (about 12-15%) has been deforested up to the present. However, there are some disturbing decreases in rainfall in other parts of the world that look like they may be a product of deforestation. For example, in Thailand there has been a drying trend during the last 30-40 years, with a 30% decrease in September precipitation (it is now 100 mm lower). Climate-vegetation models suggest that this trend could be caused by the extensive deforestation that has occurrcd in Indo-China since the 1950s. Partly, the cause of this change in climate is less recycling of rainfall by evaporation from the forest canopy (so water instead runs off as streams and rivers to the sea). But, there is also another mechanism. Drier air docs not promote as much convection in the atmosphere, because moisture does not condense out and release latent heat that might keep the air rising. Because of this there is less of the atmospheric instability needed to give rain. Interestingly, the drying trend in Thailand has been limited to a precise time of year; it is not present, for example, when the summer monsoon is in play during July and August. Modeling the system explains why this is so, and further implicates deforestation as the cause. During the monsoon, the regional influence of the evaporation from forest canopics is in effect flushed away by strong winds from the west that carry moisture in off the ocean. So. we only see the effect of deforestation once the system "calms down" after the monsoon, when the westerly wind has stopped blowing and rainfall comes mostly from more local sources of evaporation.

Another area where deforestation may have caused a long-term decrease in rainfall is southwestern Australia. Since the mid-20th century, rainfall in the area around Perth has dccrcascd drastically. River inflows arc about 42% less than they were previously, causing some major problems for the city of Perth that uses the rivers as a source of drinking water. A modeling attempt suggests that as much as half of this reduction is due to land cover change, with forests being replaced by croplands and pasture. In this case, apparently the main mechanism at work is the reduced roughness of the land surfacc. When trees were there, the uneven canopy produced greater vertical air movement, and less horizontal movement (due to the drag from the canopy). This used to favor rain production from the water vapor that blew inland from the sea or evaporated from the forest, falling back over the same forested area. Now, according to the model, moist air simply moves further inland and drops its rain there, out of reach of the catchments for Perth.

Since the 1940s, Costa Rica has lost much of its forest cover, and this seems to have changed the climate of adjacent mountains. Clouds now form less frequently and rainfall has decreased over the "cloud forest" zone in the mountains. The clouds also seem to form higher in the atmosphere, so they "miss" the mountain tops that they previously used to shroud and keep moist. The observations are backed up a model put together by Richard Lavvton and colleagues, which seems to firmly link cause and effect: less evaporation, less convection and less turbulence over the area that was once forest has changed the distribution of clouds and rain formation up in the mountains. Clouds are predicted to form less frequently, and higher in the sky when they do form. The drying of the cloud forest is thought to have been a contributing cause in the mysterious extinction of several species of colorful tree frogs-known as harlequin frogs—that only occurred in these mountains.

The effects of removing tree cover do not just occur in areas that were previously completely forested. Generally, from the models it seems that climatic effects of adding more leaf cover "saturate" at high values; there is not much difference between a very dense and a fairly dense forest canopy in terms of what it docs to climate. But, the difference between no trees at all and just a scattered open covering of trees can be far more important. Even the removal of a very open, incomplete tree cover may affect local climate. A modeling study of savannas in Brazil suggested that loss of just a fraction of the tree cover is enough to significantly decrease precipitation, increase temperature, increase wind speed and lower humidity. All these changes would tend to promote the spread of fires, further reducing the tree cover; the effect of an initial removal of trees becomes amplified so that more trees are lost.

In some areas, however, breaking up a forest cover a little might actually increase rainfall. In a model study by Roger Pielke and colleagues, replacing the original forest cover of the northern part of Georgia (in the southeastern USA) with the present mixture of fields and forest actually increased the rainfall. The key in this ease was that the forest areas were so wet that most of the sun's energy reaching them went into latent heat, not warming of the air. With so little warming occurring, it was difficult to sustain much convection (rising air). By contrast, in the field areas there was not much latent heat uptake, so the air above the field could heat up. This caused it to rise up into the atmosphere through convection, and when it did it sucked in moist air evaporating from the adjacent forest areas. This gave storm clouds and rain when it reached high enough into the atmosphere. The key here then is that the open fields act as "focal points" for convection, and the increased convection promotes rainfall.

It is important to bear in mind that in these last two studies—as with most work on regional vegetation-climate effects—there are no relevant climate data to show whether nature really works like the model says it should. In most places in the world, the carefully controlled observations of climate or vegetation cover (that one needs to test a vegetation climate feedback model by comparing the "before" and "after" scenarios) are missing. Often, this is because the "before" happened a long time ago before detailed records were being kept, or because the "after" has not yet occurred. Studies such as those in Costa Rica, Thailand and western Australia are precious gems to the world of modeling vegetation-climate feedbacks, for these provide detailed observations of climate change that correlate with a change in vegetation, and which can also be well explained by a model. Bolstered by the assurance that at least some things are well understood, modelers have the confidence to try predicting vegetation climate feedbacks on a much broader scale.


At present, while some areas of the world arc being deforested, other places are regaining the forest cover they once had. Sometimes this is due to deliberate planning and planting, as in the large areas of China and South Korea that have been re-afforested in recent decades. Other times, the return of forests is simply a by-product of social and economic forces. In the eastern USA. forests came back as a result of the collapse of the farming economy in competition with the more productive lands of the Great Plains. For example, after going down to less than 25% forest cover, the state of Rhode Island is now back to being more than 60% forested. The Low Countries of northeastern Europe have several times more forest now than they did 150 years ago. A similar proccss of abandonment of land and return of forest is occurring in parts of Eastern Europe at present. It would be interesting to know what these changes might be doing to climate, and the models intended to understand past deforestation effects can also be applied to the reverse process. Presumably, there would be a warming influence on climate as a result of the return of these temperate forests.

Continue reading here: The Role Of Forest Feedback In Broad Swings In Climate

Was this article helpful?

0 0