Box 52 Simulating climate GCMs and mesoscale models

What changes can we expect for the Earth's climate over the coming decades, as greenhouse gases increase? Because of the importance of knowing the answer to this question, a lot of effort is going in to understanding and forecasting climate change. The world's most powerful computers (known as supercomputers) are used to calculate the effects of a given rise in greenhouse gases on global climate, using a "model": a simplified world inside the computer, complete with oceans, continents, mountain ranges and an atmosphere.

Such models are also being used to investigate how vegetation creates its own climate, and what will happen to global climate if the vegetation cover is altered. As well as looking into the future, models can be made to look backwards in time, to understand how climates in the past worked, including, for example, the effects of past vegetation changes feeding back on climate.

To get a broad global perspective, climate scientists try to simulate the circulation system of the whole planet, with what is known as a general circulation model (or GCM). To model the entire global climate system is of course no easy task, and one which has taken a long time to get more or less right. Basically, the world in the computer is divided up into a grid covering its surface, and each grid cell is labeled as "ocean" or "land". If it is land, that surface grid cell is assigned an altitude, and also some attributes that relate to vegetation cover such as albedo and roughness. Up above the surface of each grid square, the atmosphere is represented as a stack of cubes. Each cube has its own composition and density of gases, and it exchanges energy with the cubes next to, above and below it, or (if it is at the bottom of the atmosphere) with the surface below it. Air is also exchanged sideways, and upwards and downwards from each grid cell, simulating the wind and also the process of convection. In the newest models, the ocean is also divided into stacks of cubes, much like the atmosphere except that these are under the surface and the fluid that fills them is not air but water. Heat and water move between these ocean boxes, simulating surface currents plus the sinking or upwel-ling of water. Winds and ocean currents push against one another, churning endlessly across the surface of the planet.

It is remarkable how manv details of the climate svstem these GCMs can simulate.

When a GCM is set up to run with the present-day atmospheric composition, the major wind belts and ocean surface currents can all be simulated quite accurately. Air masses form and move across the surface, colliding to give weather fronts. The patterns of average temperature and rainfall are closely similar to what we observe in the present world, and they go through their correct seasonal cycles. Furthermore, from year to year the global climate also goes through internally generated climate fluctuations that mimic those on the real earth.

When climate modelers are satisfied that their model works well for the present world, they can begin to tweak certain aspects of it to see how these will change the climate. For instance, they can add more greenhouse gases and observe the heat balance, rainfall and circulation systems changing in response. They can also change the vegetation cover and see how climate responds to this alteration in albedo, roughness and évapotranspiration. Some of the broader scale studies of vegetation climate feedbacks use this sort of approach to reach their conclusions about the importance of vegetation cover in making climate.

Climate modeling has come a long way in the past couple of decades, as the quantity of data that computers are able to handle per unit time has increased enormously. But, there is always still room for improvement in models. A major problem in simulating the climate is still the coarseness of models it is not possible to include every small bump or valley in the landscape, and yet such little microclimatic differences might add up to significant broad-scale effects. Many-processes such as the formation of thunderclouds occur at a scale smaller than a single grid cell so they must be assumed to occur rather than simulated directly. Decreasing the size of the grid cells in the model increases its accuracy, but doing so magnifies the computing task enormously. Modelers have to strike a balance between the time taken to run a model, and the accuracy that it can produce. As computing power has increased, the size of the grid cells in models has decreased from 5 x 5° in the 1980s, to 0.5 x 0.5° in the latest models. At the time of writing, the world's most powerful computer system located in Japan was built especially for the task of running the most sophisticated climate models.

One way to get more fine-scale accuracy, without running up against enormous computing problems, is to focus on simulating one area of interest in detail and leaving the rest of the world outside it at a lower resolution. For this purpose, climate modelers use a special add-on model that works at a regional scale, known as a mesoscale model. A mesoscale model works in many ways like a GCM except that its grid squares and boxes are smaller and cover a more restricted area the region simulated by such a model will be at most a few hundred or a thousand kilometers across. A mesoscale model partly creates its own climate from the sunlight that falls on it, but it slots into a broader GCM which supplies heat and water vapor in at the edges, and takes these away from the edges too when the wind blows out from the area. A mesoscale model is ideal for exploring how detailed changes in vegetation will affect a regional climate. Many of the vegetation climate feedback studies mentioned in this book were carried out with mesoscale models coupled to broader GCMs.


There may be some places in the world where "climate-engineering'' by humans altering vegetation cover has already occurred, albeit unintentionally. In the 1800s the grasslands of the central USA were transformed at a pace and on a scale unmatched in any other region in history. Settlers poured westwards in their millions, ploughing up the deep prairie soils to plant wheat and corn fields stretching for hundreds of miles.

Did this affect the climate? The debate about it goes back a long way, to the time when the land was still being ploughed up. In the Plains climate, a drier than average year could prove disastrous for crops, so there was plenty of interest in ensuring that the rainfall was as reliable and abundant as possible. In the 1880s. Samuel Aughey, a professor in the young state of Nebraska, suggested that ploughing a prairie soil helps it to retain water better because with the mat of vegetation on the surface broken water soaks in rather than running straight off into rivers. This store of water held in the crumbly soil will then evaporate, being recycled as rain which falls to earth again, instead of being lost to rivers and the sea. This idea became encapsulated by the plains farmer's adage: "Rain follows the plough/* Although the idea got a lot of attention, it is now thought that ploughing actually docs not have such an cffect on rainfall.

Others at the time suggested that the best thing to do to ensure steady rainfall was to plant more trees. In the 1860s a US government official named Joseph Wilson pointed out that since deforestation seemed to have decreased rainfall in other parts of the world (sec Chapter 6), planting trees in the Great Plains would surely increase the rainfall there. He advocated covering a third of the Great Plains in trees to ensure an adequate supply of rain. Congress was impressed enough by his arguments to pass an act that offered free land parcels to fanners who planted a certain percentage of their land with trees. However, the farmers were not motivated by these incentives few trees were planted, and the act was eventually repealed.

More recently, aided by modern climatological knowledge and computers, scientists have been able to take a more informed look at the effects of converting the prairies to grain fields. Some modeling studies by Eastman and colleagues suggest that replacing the grasslands of the central Plains with crops caused the peak temperature reached during the afternoon to increase by between 1 and 6 C, depending on the location and time of year. The warming in the model strengthens during the growing season, and decreases as the crops are harvested. The most important factor in causing this warming is that the crops have fewer leaves per unit area than the grasslands. With fewer leaves there is less transpiration of water, and less uptake of energy in latent heat; hence, the air can get warmer over the crops.

Settlers may have affected the climate across the Great Plains even before they had managed to plough up most of the land for crops. Up until the mid-1800s, the Plains supported vast herds of bison, numbering in the tens of millions. The mass slaughter of these animals during the early phases of settlement would have greatly reduced the grazing of prairie grasses. With more leaf area accumulating uneaten, there would have been more evaporation of water from the leaves. Climate models suggest that this could have cooled summer temperatures by 0.4-0.8 due to extra latent heat uptake by the evaporating water. This would then have been followed by the main phase when the farmers ploughed the landscape and planted crops, which reduced leaf area to below what it had been in the grazed prairie and caused a raising of temperature as explained above.

In the modern Great Plains, particularly towards the western edge, farmers irrigate their crops with water from underground aquifers. What does all this extra

Fort \ Collins

Figure 5.5. Temperature map for a warm day in northeastern Colorado. Irrigated areas such as suburbs and agricultural land have cooler temperatures than non-irrigated areas. Surface temperature at 13:00, 1 August to 15 August 1986. Contour from 38 to 28 by 2. After: Bonan.

water on the fields do to the climate? Modeling studies suggest that the uptake of heat into evaporation from irrigated crops (compared with non-irrigated crops or prairie) will cool the air and create a sort of "sea breeze" blowing outwards to nearby hotter, non-irrigated areas. Measurements comparing irrigated and non-irrigated areas of northeastern Colorado show that, as the models predict, temperatures are several degrees C cooler where there is irrigation, due to latent heat uptake, altered wind patterns and cloudiness (Figure 5.5). As irrigation in the area has expanded over the last 45 years, there has also been a cooling trend in climate, as would be expected. The models also predict an increase in rainfall over irrigated areas as a result of both the extra water evaporated, and the movement of air that results from the temperature contrasts between irrigated and non-irrigated land. Observations from northern Texas show that extensively irrigated areas have more rainfall than otherwise similar areas that do not get much irrigation.

On the other side of the world, parts of another arid region may have been affected by climate feedbacks that result from land use change. In southern Israel over the last 50 years, intensification of farming (including increased irrigation), reduced grazing and tree-planting has resulted in lower albedo and more évapotranspiration from vegetation. Since the early 1960s there has been a dramatic increase in autumn rainfall, by as much as 200-300% depending on the location. It seems plausible that the climatc change has been a result of the progressive change in land use in this area. The increased upwards movement in the atmosphere above these lands seems to suck in moist air off the Mediterranean, which gives much of the rainfall.

The Sinai desert of Hgypt has cooler daytime temperatures than the adjacent Ncgcv desert of Israel, by 3.5-5C in the early afternoon. It seems that the key factor that makes the Sinai cooler is its lack of vegetation, due to a lot more goat and sheep-grazing and cutting of firewood. With more high-albedo soil exposed, the Sinai reflects back more sunlight and cannot heat up as much. But, doesn't this contradict what 1 said at the beginning of the chapter—that dark vegetated areas tend to be cooler because they evaporate more moisture? In fact, it is the exception that proves the rule that, without evaporation, dark vegetated areas would always be hotter. Conditions in the Sinai and Negev are so dry that there is no soil moisture to evaporate much of the year. So, the dark vegetation cover in the Sinai (although it is fairly sparse) merely absorbs the sun's rays but does not suck heat away into transpiration.

In slightly moistcr—but still arid—areas such as the Sonoran Desert in the southwest USA and Mexico, adding a bit more vegetation makes things cooler not hotter. The heavily grazed Mexican side of the border is several degrees hotter during the day than the lightly grazed US side. This is because in this case there is enough moisture in the soil for the extra leaves on the US side to have a cooling effect by transpiring more water, and this dominates over the warming caused by the darker vegetated surface.

Continue reading here: Box 53 Interactive vegetation schemes in dimate modeling

Was this article helpful?

0 0