本文分别利用最小二乘拟合迭代法、牛顿迭代法、TVM法和Martano法计算2003年长白山森林地表粗糙度。结果表明,Martano法计算结果偏小,其他三种方法结果基本一致。不确定性分析表明,Martano法的不确定性与参与计算的数据量有关。拟合迭代法和牛顿迭代法的不确定性在于,通常假设u_*不随观测高度而变化,而实际u_*随观测高度变化。若相邻两个观测高度之间的u_*变化1％,地表粗糙度计算误差8.3%。TVM法由于参数C₁的选取范围为0.9~1.05,引起地表粗糙度不确定性平均值为29.9%。经计算得到长白山森林在各种大气层结条件下的地表粗糙度,拟合迭代法得到的2003年稳定、中性和不稳定条件下的地表粗糙度年平均值分别为2.642、2.103和1.616m。

There are four typical surface roughness length calculation methods, which are iterative method by least-square fit, Newton iterative method, Temperature Variance Method (TVM method) and a method proposed by Martano. The former two methods need wind speed and temperature profiles in several levels, while the latter two methods use three-dimensional sonic anemometer data in a single level. In this paper, the surface roughness length in various atmospheric stratifications over forest underlying in Changbai Mountains Experimental Station is calculated by the four methods respectively, with wind speed and temperature profiles and three-dimensional sonic anemometer data in 2003. Discrepancy among the four methods and uncertainty of each method are analyzed. The results indicate that, except Martano method, there is minor difference between surface roughness lengths calculated by the other three methods, while roughness length calculated by Martano method is obviously much smaller. By uncertainty analysis of Martano method, it is found out that smaller roughness length calculated by Martano method is probably related to data amount, and calculation results change obviously with data amount. Otherwise, by making uncertainty analysis for the other three methods, it is found out that as for iterative method by least-square fit and Newton iterative method, calculation roughness length is influenced by friction velocity u_*, and if u_* changes 1% between two neighborhood measurement heights, the calculated roughness length would change by 8.3%. As for TVM method, parameter of C₁ changes in such a certain range from 0.9 to 1.05 that the calculated roughness length changes by about 29.9%. By calculating surface roughness length in various atmospheric stratifications with four methods, it is verified that surface roughness length changes with atmospheric stratifications. Surface roughness length in unstable stratification is the smallest, the one in stable stratification is the highest, and in the neutral stratification is medium. Also in this paper specific values in various stratifications are calculated, The results are 2.642, 2.103 and 1.616m for stable, neutral and unstable stratification roughness length, respectively.