초록
온도기반 생태모형을 경관규모에 적용하기 위해 널리 쓰이는 BioSIM을 우리나라 환경에 도입할 경우 예상되는 문제점을 파악하기 위해 먼저 일 최고기온 추정과정을 검토하였다. 과열지수 대신 사용되는 개방도의 신뢰성을 검증하기 위해 전라북도 순창군 동계면 전역을 대상으로 10${\times}$10m 면적 단위로 총 57만여 개 지점의 개방도를 계산하고, 같은 지점에 대해 추분, 하지, 동짓날의 과열지수를 계산하였다. 각 날짜별 과열지수의 변이를 개방도에 의해 설명하기 위한 2차 회귀식을 작성한 바 회귀식의 결정계수($R^2$)는 동지에 0.44, 하지에 0.50, 추분에 0.56으로 Regniere(1996)의 추정치 0.7-0.9에 비해 크게 낮았다. 따라서 개방도를 사용하여 추정된 복잡지형의 일 최고기온값은 신뢰도가 낮아 생태모형의 구동변수로 부적합하므로 반드시 과열지수를 직접 계산해서 사용해야 한다.
'||'||'||'&'||'||'||'quot;Overheating index'||'||'||'&'||'||'||'quot;, the normalized difference in incident solar energy between a target surface and a level surface, is helpful in estimating the spatial variation in daily maximum temperature at the landscape scale. It can be computed as the ratio of the 4-hour cumulative solar irradiance surplus or deficit from that over a level surface to the maximum possible deviation (15 MJ $m^{-2}$ ) during the midafternoon. Ecosystem models may, for simplicity, use an empirical proxy (exposure index) variable combining slope and aspect in place of the overheating index to account for the variation of midafternoon solar irradiance. A comparative study with real-world landscape data was carried out to evaluate the performance of exposure index in replacing the overheating index. Overheating indices for summer solstice, fall equinox and winter solstice were calculated at 573,650 grid cells constituting the land surface of Donggye-Myun, Sunchang County in Korea, based on a 10-m DEM. Exposure index was also calculated for the same area and fitted for the variation of overheating index to derive a 2$^{nd}$ -order linear regression equation. The coefficient of determination ($R^2$) was 0.50 on summer solstice, 0.56 on fall equinox, and 0.44 on winter solstice, respectively. These are much lower than the theoretically calculated $R^2$ values ranging from 0.7 in summer to 0.9 in autumn. According to our study, exposure index failed to accurately predict the cumulative solar irradiance over a complex terrain, hindering its application to daily maximum temperature estimation. We suggest direct calculation of the overheating index in preference to using the exposure index.
