Variations in surface roughness of heterogeneous surfaces in the Nagqu area of the Tibetan Plateau

Temporal and spatial variations of the surface aerodynamic roughness lengths (Z(0m)) in the Nagqu area of the northern Tibetan Plateau were analysed in 2008, 2010 and 2012 using MODIS satellite data and in situ atmospheric turbulence observations. Surface aerodynamic roughness lengths were calculated from turbulent observations by a single-height ultrasonic anemometer and retrieved by the Massman model. The results showed that Z(0m) has an apparent characteristic of seasonal variation. From February to August, Z(0m) increased with snow ablation and vegetation growth, and the maximum value reached 4-5 cm at the BJ site. From September to February, Z(0m) gradually decreased and reached its minimum values of about 1-2 cm. Snowfall in abnormal years was the main reason for the significantly lower Z(0m) compared with that in normal conditions. The underlying surface can be divided into four categories according to the different values of Z(0m): snow and ice, sparse grassland, lush grassland and town. Among them, lush grassland and sparse grassland accounted for 62.49% and 33.74 %, and they have an annual variation of Z(0m) between 1-4 and 2-6 cm, respectively. The two methods were positively correlated, and the retrieved values were lower than the measured results due to the heterogeneity of the underlying surface. These results are substituted into the Noah-MP (multi-parameterisation) model to replace the original parameter design numerical simulation experiment. After replacing the model surface roughness, the sensible heat flux and latent heat flux were simulated with a better diurnal dynamics.

Automated summary

Temporal and spatial variations of surface aerodynamic roughness lengths (Z(0m)) in the Nagqu area of the northern Tibetan Plateau were analyzed using MODIS satellite data and in situ atmospheric turbulence observations. The results showed a seasonal variation of Z(0m), with an increase from February to August and a decrease from September to February, influenced by snow ablation and abnormal snowfall years. The underlying surface could be divided into different categories based on Z(0m) values, with lush grassland and sparse grassland accounting for the majority of the area and having annual variations of Z(0m) between 1-4 and 2-6 cm, respectively.