Open AccessImportance of land use versus atmospheric information verified from cloud simulations from a frontier region in Costa Rica
Author(s) -
Ray Deepak K.,
Pielke Roger A.,
Nair Udaysankar S.,
Welch Ronald M.,
Lawton Robert O.
Publication year - 2009
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.67
H-Index - 298
eISSN - 2156-2202
pISSN - 0148-0227
DOI - 10.1029/2007jd009565
Subject(s) - land cover , baseline (sea) , cloud cover , climate model , climatology , precipitation , environmental science , frontier , meteorology , climate change , scale (ratio) , land use , cloud computing , geography , computer science , geology , cartography , archaeology , operating system , oceanography , civil engineering , engineering
Land‐use/land‐cover (LULC) change has been recognized as a key component in global climate change, and numerous climate modeling studies at regional to global scales document this. The research strategies have invariably been to first conduct baseline simulations of current conditions to evaluate model performance. Then simulation of regional climate with land cover changes (LCC) implemented within the model allows differences with the baseline simulation to be used as evidence of global to regional‐scale climate impacts of LCC. However, even state‐of‐the‐art regional climate models require two data sets to conduct reasonable baseline simulations. These are representative current land cover and atmospheric information over the study region. In frontier and developing areas (where most of the rapid land‐use conversion is taking place), these data sets are frequently unavailable and the errors in simulations are due to either inaccurate land cover, insufficient atmospheric information, nonrepresentative model physics, or a combination of one or more of the above. This study shows that in one frontier region, that surrounding the Cordillera de Tilarán of Costa Rica, the accuracy of simulating clouds decreases by 1% to 3% if default model land cover information is used. If the atmospheric data sets used are the ones usually available to researchers (with land cover information held constant), then the model accuracy is reduced by 21% to 25%. Model runs without updated land cover or atmospheric information reduce model accuracy slightly further. Precipitation comparisons also provide similar results. This study thus shows that the critically important data set for conducting accurate simulations is not land cover information but atmospheric information. Researchers may similarly get significant increase in the accuracy of their baseline simulations elsewhere by using radiosondes/rawinsondes over their study region. Finally, since atmospheric information is not available for different landscape scenarios, assessments of the relative role of LULC change will have to continue to rely on using the standard atmospheric data set and the acceptance that the use of more detailed atmospheric data to initialize and provide lateral boundary conditions would have reduced the uncertainties in such landscape sensitivity studies.