EdGCM: Climate Modeling for Research and Education - A Climate Model Vacation Planner

A Climate Model Vacation Planner

Outreach - Showcase
Written by Ken Mankoff
Friday, 12 May 2006
Abstract The EdGCM tagline is "EdGCM: Climate Modeling for Research and Education". In this study the tagline was changed to "EdGCM: Climate Modeling to Plan my Vacation". While planning a vacation to Cambodia last fall, I decided to use EdGCM to verify information provided in the guidebooks, and then discovered it provides much more... Cambodia, located at 13^o00" N, 105^o00" E occupies an area of 176,520 sq km [1,4]. The grid cell containing Cambodia is approximately 5 times as large as the country itself. Statistically speaking, using EdGCM to look at the climate for one month in one cell would be considered a questionable practice, but assuming that it provides nothing of value is incorrect - the results are both in agreement with the qualitative information provided by the CIA World Factbook [1] and most guidebooks [2,3]. Guidebooks provide temperature and precipitation only, while EdGCM provides other valuable vacation variables such as cloud cover (tanning), soil moisture and ground wetness (camping), wind speed (windsurfing), and diurnal temperature variation. Introduction Figure 1: Annual precipitation EdGCM is a coarse resolution global climate model (GCM) with a grid spacing of 8^o latitude by 10^o longitude. Global in this context does not only mean that EdGCM models the entire globe, but that the outputs should be viewed in a global context for best accuracy. EdGCM can be used for quantitative regional studies when regional is defined as 'continent size'. That means EdGCM provides the best results when looking at results with spatial resolutions that cover continents, hemispheres, or the entire planet. Similarly, the temporal resolution that most scientists examine with climate models are long-term averages. Standard output for most GCMs is the monthly average, although seasonal and annual averages are commonly used as well. I used EdGCM to look at 1 cell for 1 month. Figure 2: December surface air temperature Initial Google results return the CIA World Factbook page [1] where it states that November and December are the coolest and driest months, and therefore the best months to visit. This was verified by running the Modern_PredictedSST run in EdGCM and using Analyze Output and EVA to process and look at maps of the grid cells around South East Asia for a sequence of months using the Precipitation and Surface Air Temperature variables. EdGCM and the CIA World Factbook appear to agree with each other. Further searching both online and at a local bookstore showed that guidebooks appear to just copy the information found on the CIA pages. Regional Maps Figure 3: December minus annual temperature difference Figure 1 shows that the GCM simulates Cambodia's annual average precipitation as being 8 to 9 mm/day. This is significantly higher than the global average of 3 mm/day, though somewhat less than the global maximum of 13 mm/day. Figure 2 shows a simulated December average temperature of 20 to 22 degrees C. Flipping from month to month in EVA (EdGCM's visualization software) can be used to determine which months are coolest and/or driest, but this is not an optimal method of visualizing the data. Differencing the data, in this case December - Annual averages, provides far more insight, revealing (Figure 3) that December temperatures are 3 to 7 degrees cooler than the annual average. Local Plots Figure 4: Regional climate line plot of temperature, precipitation, and cloud cover Latitude longitude maps provide a good way to view global and regional scale results, and differencing maps gives extra insight into the data. But since the region of interest is only one cell, most of the map is wasted space. Visualizing the data outside of EdGCM and EVA allows a time series of multiple variables in only the region of interest. Figure 4 shows precipitation, surface air temperature (average, min, max), and cloud cover for the cell containing Cambodia (x=28, y=13) of the GCM. The X axis represents months (January through December) while the Y axes are temperature and precipitation (together), and cloud cover respectively. The dashed green line is the yearly average precipitation and the vertical dashed gray line separates November and December (the approximate time of my vacation). Figure 4 shows that December and January have the coolest temperatures (low, high, and average). The diurnal variation is greatest in January and the maximum daytime temperatures occur in May while the maximum nighttime temperatures are in June and July. The rainy or monsoon season is May through September, with the dry season having the same average from month to month. The cloud cover peaks during the monsoon season, and the maximum sunlight occurs in November. The months of interest for the vacation are therefore November and December. While December might be slightly cooler, there is significantly more cloud cover. In Situ Experience I spent November 18 through December 4 in Cambodia. The first half was inland in the northern section at Angkor Wat. The 2nd half was on a small island in the Gulf of Thailand. While I did not travel with a thermometer, or any scientific instrumentation with which to verify my analysis and selection of dates, it was hot during the day, comfortable to slightly chilly at night, and sunny. There were occasional late afternoon rain showers while on the southern coast. When in the north there was often a haze, not something characterizable as cloud cover. Limitations and Further Investigation The obvious limitations of using a GCM to view one cell for one month do not appear to hurt this analysis based on the agreement between the quantitative GCM outputs and the qualitative response from [1,2 and 3], although there are still dangers to planning a vacation this way. For example, local streams will affect soil moisture and ground wetness. Local mountains, water, etc. will effect precipitation, low level cloud cover and possibly even mid level cloud cover . High level clouds will be less affected by surface variability. Local cloud cover will change local precipitation. Coast versus inland will have significant effects on surface air temperature. Figure 5: Long term standing water. Is it handled in the GCM? The results of this type of "single cell" analysis would clearly be more trustworthy if we examined the GCM's climate variability for the grid cell over Cambodia. This could either be accomplished by looking at the results for numerous years from a multi-year simulation, or by running an ensemble of experiments, each with slightly altered initial conditions (change the Random No. Seed in Setup Simulations). When planning a camping trip it would be nice to know which of the myriad of climate variables that one can view using EdGCM are most relevant. For example, the difference between "soil moisture" and "ground wetness" isn't clear. Throughout Southeast Asia there are significant areas of standing water (rice paddies, etc.). During the wet months large portions of the country are inundated with water. Figure 5 is an aerial photo that shows the Mekong river winding outside of Phnom Phen, and standing water. Visual estimates of this photograph put standing water at 50% to 75%. Does the GCM handle this? Is it a temporary (6 month) inland sea? Is soil moisture set to 100% for this period of time? 100% soil moisture is not an accurate representation of a large inland lake, as the properties and physics of water are significantly different than land, even land with saturated soil. Conclusion EdGCM, and therefore presumably GCMs in general, are able to do local scale studies when the scientific accuracy of the results does not require a high degree of quantitative precision. EdGCM provides similar information (spatially and temporally) as guide books and the CIA World Factbook for the variables they cover, and then provides many additional variables. Also, I conclude Cambodia is a nice place. Bibliography [1] http://www.cia.gov/cia/publications/factbook/geos/cb.html [2] http://www.tourismcambodia.com/Highlights/be4ugo/Climate.asp [3] http://www.lonelyplanet.com/worldguide/destinations/asia/cambodia/essential?a=weather [4] http://en.wikipedia.org/wiki/Cambodia Discussion Discuss this article...