Micro-climate Solar Modeling over Complex Terrain
Track: Climate, Weather and Atmosphere
Author(s): Nathaniel Vandal, Bill Hegman

Incoming solar radiation (insolation) is the single most important source of energy at the Earth's surface. Solar energy variation on the landscape is the primary determinant for the spatial distribution of almost all natural and biological processes, including human life. There is, therefore, tremendous importance in the ability to accurately model incoming radiation on the landscape. This paper analyzes the results of Environmental Systems Research Incorporated's (Esri) Solar Analyst on complex topography that includes land forms and buildings. Little solar modeling has been conducted on this scale, and no validated results have been published to date. The analysis in this paper starts with the development of high a resolution complex DEM of the Middlebury College campus which has a topographical range of 285 feet and contains over 80 buildings. Radiation was calculated over the study area with atmospheric data available from the National Solar Radiation Database. Validation of the model was conducted using on-site solar pyronometers placed at strategic locations within the study area. In addition to the validation study, a sensitivity analysis of the user specified input parameters was conducted to help determine their effects on the accuracy of the model. Validation results revealed an average error of 11%. Of the potential sources for error identified in this study, snow events were determined to be the most significant with an average of 7% of the study days having snow events. These results are magnified by the short time interval of the study, highlighting the need for longer measurement durations in future studies. Results from the validation study helped to determine a hierarchy for model input parameter selections, which was as follows: hour interval, day interval, skysize, and calculation directions. The effects of zenith and azimuth divisions were not significant. Input parameter settings were chosen from the sensitivity analysis to achieve 1% convergence accuracy (day interval 1, hour interval 0.1, skysize 1000, calc. dirs. 48) and were used to create radiation maps for the entire study area. Radiation patterns in the area modeling results matched up with empirical interpretations from topography, with the exception of some abnormal ground shading on the south side of buildings, which requires additional validation before being determined erroneous. In conclusion, the results from this study highlight the need for enhanced DEM creation, and increased modeling of solar radiation on complex terrain.

Nathaniel Vandal
Middlebury College
MC Box2004
Middlebury College
Middlebury , VT 05753
Phone: 802-272-8330
E-mail: nvandal@middlebury.edu

Bill Hegman
Middlebury College
95 Camels Hump Rd
Huntington , VT 05462
Phone: 802 434-4759
E-mail: bhegman@middlebury.edu