|
Projections of Future Climate from the GCMs |
|
| Peter J. Sousounis, Meteorologist, Michigan State University | |
|
Intricacies of the General Circulation Models (GCMs) that we used for the Great Lakes Regional Assessment The Great Lakes Regional Assessment as well as the other Regional Assessments that participated in the national assessment used output from two General Circulation Models, the Canadian Model and the Hadley Model. These were some of the most sophisticated models available at the time. The reason for this is that the parameterization of the different processes that occur in the atmosphere as well as in the ocean were incorporated to the best of anyone´s knowledge and ability. Additionally, the way in which carbon dioxide was increased was much more realistic. First generation models instantaneously double the amount of CO2 in the atmosphere and then waited for the models to come to equilibrium. These models provide much more realistic scenarios in that carbon dioxide was increased incrementally, 1 percent per year, compounded annually through the end of this century. “Climate Scenarios —Temperature Changes” shows projections of temperature change for 4 different regions. Temperature change, according to the Canadian Model, is rather significant, on the order of about 5 degrees Celsius by the end of this century. The Hadley Model, is more modest, but clearly shows some increase in temperature as well, on the order of about 2-3 degrees Centigrade. Temperature is one aspect that we, as climate scientists, feel very confident will occur. Thickness (The thickness of the lower half of the atmosphere) Just as you heat a column of air the column becomes taller, so does the atmosphere. The thickness of the lower half of the atmosphere is related to the surface temperature. That is, the greater the surface temperature, the greater the thickness will be. Thickness is being used here, as opposed to surface temperature, because it is much more reliable. General Circulation Models (GCMs) and even the short term forecast models that are used are to do the day-to-day forecasts do a much better job at forecasting large-scale parameters, air motions that are occurring several kilometers above the surface, than they are at forecasting what´s actually happening at the surface. The Canadian Model shows increases in thickness for all months, more or less uniformly, but some of the most dramatic increases are occurring in winter and we expect that the temperatures in winter will increase more than they will in the summer. The Hadley model shows overall smaller changes than the Canadian Model. Extreme Temperature Days These factors may be more relevant from an agricultural perspective and from a human health perspective in terms of heat morbidity and heat mortality. The chart of “Interannual Variability in Hot Days for SW Michigan” indicates the number of 90 degree plus days per year. At present between 10-20 such days occur in SW lower Michigan and that agrees very will with observations in Detroit. The number of 90 degree plus days the models are projecting by the end of the century are shocking. The Canadian Model is showing that these days may occupy the better or entire part of the summer. The Hadley Model is a bit more modest, still showing about half of the days during the summer will be 90 degrees or higher. There is quite a range, between 40-90, but what the results show is that there will be a considerable increase. Because of this considerable increase, it follows that the year-to-year variability will increase. If you only have on average 10 days, then it is hard to have a tremendous year-to-year variation, but when you have 40 days, you have much greater year-to-year variation. The Hadley model shows that there may be some years that are not too different from what we are experiencing at present, but there may be some years where it is considerably hot. (Changes in growing season) If the mean temperature is increasing, that will translate to earlier planting dates and later harvest dates; the general growing season will increase. What we don´t know is how the extreme events factor into this. Currently the Great Lakes Region receives just under 1000 mm of precipitation in the Canadian Model is showing on average that there will be an increase in precipitation. The Hadley model shows an even greater increase in precipitation. Month by month breakdown of precipitation from the two models, the Canadian model shows changes in precipitation run from 1247 mm to 1336 mm about an 8% increase. The Hadley Model shows 957 mm observed to a future projection of 1075 mm, about a 12% increase. The Canadian model shows that most of that increase occurs during the first 6 months of the year—which translate to preparing the soil in terms of soil moisture for planting for agricultural purposes. The Hadley model shows that most of the increase is occurring during the second half of the year. Why should precipitation increase? This is a question I continue to explore. It is not clear whether the precipitation increases will come from more frequent storms, slower storms, stronger storms, wetter storms, or more efficient storms. Breakdown of participation increases from intensity perspective, “Annual Precipitation Category Changes for Detroit, MI.” Results for the Canadian and Hadley Models, breakdown in precipitation events: In the Canadian Model, the first three are “light events” up to 1/2 inch (12.7 mm), the last two categories are “heavy events” greater than 1/2 inch. The number of days corresponding to light events decreases and the number of days with heavy events increases, but not enough to compensate. There is a net decrease in number of days with precipitation by the end of this century. The amount of precipitation associated with these events increases considerably. A decrease of 17 mm owing to the decrease of light events, but an increase of 155 mm per year owing to the increase of heavy precipitation events. The Hadley Model shows something very similar. A significant decrease in the number of light events, and a smaller increase in the number of heavy events. The amount of precipitation contributed by these events shows a considerable increase from the heavy events. Interannual variability is a key concern. There is a lot of year-to-year variability in the precipitation amounts. On average the year-to-year variability of heavy precipitation days will increase by the end of this century. The amount of precipitation from these heavy events will also increase. For example in the Canadian Model one year the heavy events might contribute 900 mm to the total precipitation amount, the next year they might contribute 1400 mm. Reasons for heavy precipitation increases (From the Canadian Model, Hadley is similar) There is a low pressure in southern Wisconsin for the current scenario corresponding to heavy precipitation events, and for the future climate scenario, these results suggest that the airflow, the synoptic situation, that is responsible for these heavy precipitation events may not change much. This suggests that the storms may just be wetter storms owing to the fact that temperatures are higher and the saturation vapor pressure is higher. This accelerates the water in the precipitation cycle. The change in the number of cyclones projected by both models also decreases. There are differences in which months those decreases occur. Hadley model shows a tremendous decrease in the first half of the year and this is the portion of the year where climatologically we have the greatest number of storms. Greatest number of storms that move through our Region are in April. Winds in general, both models show a net decrease in overall wind speed. Not necessarily greater in any one month, than in another, but for example, the Hadley Model showing a drop from about 8.72 meters/sec to 8.56 meter/sec; the Canadian from 8.83 meters/sec to 8.17 meters/sec. Not tremendous decreases, but they may be significant when we look at these on a year-to-year or day-to-day basis. This hasn´t been done yet. The mean upper airflow conditions for winter for current and future scenarios, from the two models. This is at 300 and 250 millibars. The wind speeds are in excess of 120 knots. This suggests that El Nino events might become more frequent, if not more intense. From that it may follow that winter conditions and those that follow in the spring and summer might be very typical of those that occur during El Nino years. Climate change in the Great Lakes region will be manifested by changes in winds and storm tracks as well as by changes in temperature and precipitation. We need to learn more about the year-to-year variations as well as the day-to-day variations. We live on a day-to-day basis, not in a 20-year mean sense. From the results it looks like extremely hot days will become more frequent and extreme heavy precipitation events will also become more frequent. There are also some indications that interannual variability will increase. That has to be examined more closely. Q 1: John Oakley, When youÕre talking about wind speeds and added heating, 90 degree plus days, youÕre looking at evaporation, both plusses and minuses; and youÕre talking about a deficiency of rainfall that we have to rely on soil moisture to grow crops, are irrigation possibilities more or less valid for production agriculture? R: Peter Sousounis: I personally havenÕt looked at evaporation. Brent Lofgren, in the audience, has examined how lake levels might change evaporation over the land surfaces as well as the lake surfaces. It does follow that with hotter days, even though we expect an increase in precipitation, there will be definitely be an increase in evaporation. There is the potential for dryer soil, even though the Canadian Model shows that the heavier precipitation would be weighted more towards the first half of the year and that might prime the soil more. That might just delay the problem of the drying of the soil. R: Jeff Andresen: Looking at the same two models in the agronomic sense, we found that there is enough water to compensate for the increased rate of evaporation. In general, especially as we went towards the end of this century in these simulations, that there was less overall stress. There is a longer growing season and enough extra precipitation, resulting in lower risk and higher yields, according to the models. Q 2: Gary Heilig, YouÕve talked about the highs, but what about the low temperatures? R: Peter Sousounis It appears as if the greatest changes in temperature will be those that occur in the winter for minimum temperature. There are several reasons for that, one of which is the greater increase in cloudiness, which will keep radiational cooling from occurring as often as it may in the fall and winter months. Another reason is the overall increase in temperature. The Polar Regions, by far will increase in temperature the greatest and that is simply a result of an expectation that much of the polar ice caps will melt. That will reduce the albedo (the amount of energy that is reflected back to space will now be absorbed at the surface, leading to an increase in temperature.) The decrease in pole to equator temperature gradient essentially drives the frequency of storm development as well as overall wind speed. That is one reason that I can say with some confidence why I expect the number of cyclones, or low pressure systems to decrease by the end of the century, as well as the wind speed. Q3: Leland Townsend Is someone going to speak about the Great Lakes water levels? I understand theyÕre down substantially. R: Peter Sousounis: We had a workshop last year that focused on lake levels and water resources. The expectation is that lake levels will decrease in the mean sense. The decreased lake levels that weÕre experiencing now are not necessarily a reflection of climate change, rather they are more an indication of what we expect to occur more frequently and these shorter term variations that weÕve been experiencing. For example the record high lake levels in the 80Õs and now the much reduced lake levels 20 years after that. If we superimpose that on a lower mean lake level, then we have to entertain the possibility of record low lake levels occurring in the future. More information about the Lake Levels Workshop and the Water Ecology Workshop is available on our website, www.geo.msu.edu/glra. Q4: George Hubka: From your findings that weÕre going to be getting water more day after day and the hotter periods stringing together, will that portend that weÕre going to need fewer, but maybe longer more intense irrigation sessions? R: Jeff Andresen: The results that we would have suggest that for some crops it might even be less, given if the precipitation forecasts are right. It would depend on the crop. It would that there will be a longer season, so in some cases there may be more water use. |
