Variations of surface roughness on inhomogeneous underlying 1 surface at Nagqu Area over the Tibetan Plateau 2

470 Li J., Hong Z., Sun S., 2000. An Observational Experiment on the Atmospheric Boundary Layer in 471 Gerze Area of the Tibetan Plateau[J]. Chinese Journal of Atmospheric Sciences, 24(03): 301-312. 472 (in Chinese with English abstract) 473 Li L., Chen X., Wang Z., et al, 2010. Climate Change and Its Regional Differences over the 474 Tibetan Plateau[J]. Advances in Climate Change Research, 6(03): 181-186. (in Chinese with 475 English abstract) 476 Liu J., Zhou M., Hu Y., 2007. Discussion on the Terrain Aerodynamic Roughness[J]. Ecology and 477 Environment, 16(06): 1829-1836. (in Chinese with English abstract) 478 Ma Y., Wang J., et al, 2002. Analysis of Aerodynamic and Thermodynamic Parameters on the 479 Grassy Marshland Surface of Tibetan Plateau[J]. Peogress in Natural Science, 12(01): 36-40. (in 480 Chinese with English abstract) 481 Ma Y., Yao T., Wang J., et al, 2006. The Study on the Land Surface Heat Fluxes over 482 Heterogeneous Landscape of the Tibetan Plateau[J]. Advances in Earth Science, 21(12): 483 1215-1223. (in Chinese with English abstract) 484 Su B., Zhao M., Ren J., 1999. Influence of Scalar Roughness Lengths on the 485 Biosphere-Atmosphere Transfer[J]. Chinese Journal of Atmospheric Sciences, 23(03): 349-358. 486 (in Chinese with English abstract) 487 Sun G., 2016. The Upscaling Analysis of Surface Fluxes of Alpine Grassland over the in Northern 488 Tibetan Plateau[D]. Lanzhou: Cold and Arid Regions Environmental and Engineering Research 489 Institute Chinese Academy of Science, 1-134. (in Chinese with English abstract) 490 Wang J., 1999. Land Surface Process Experiments and Interaction Study in China-from HEIFE to 491 Imgrass and GAME-TIBET/TIPEX[J]. Plateau Meteorology, 18(03): 280-294. (in Chinese with 492 https://doi.org/10.5194/hess-2020-360 Preprint. Discussion started: 25 August 2020 c © Author(s) 2020. CC BY 4.0 License.


33
Known as the "third pole" of the earth (Jane, 2008), the Tibetan plateau has an average 34 altitude of over 4000m, accounting for a quarter of China's territory. It is located in the southwest 35 of China, adjacent to the subtropical tropics in the south and reaching the mid-latitude in the north, 36 making it the highest plateau in the world. Due to its special geographical location and 37 geomorphic characteristics, it plays an important role in the formation, outbreak, duration and 38 intensity of Asian monsoon especially in the global climate system. (Yang et al., 1998;Zhang et 39 al., 1998;Wu et al., 1999Wu et al., , 2004Wu et al., , 2005Ye et al, 1998;Wu et al, 1998;Tao et al, 1998). Lots of 40 research shows that (Wu et al., 2013;Wang, 1999;) the land-atmosphere interaction 41 on the Tibet plateau plays an important role in regional and global climate. Wang (Wang, 1999) 42 pointed out that high latitude areas and mountainous areas are sensitive areas of climate change, 43 especially the land-atmosphere interaction located in the plateau area of middle latitude (including 44 large areas of permafrost), which plays an extremely important role in regional climate and global 45 climate. The plateau monsoon is closely related to the intensity and location of the south Asian 46 high (Xun et al, 2002). The correlation between the dynamic index of plateau monsoon and the 47 pelagic meridional wind shows that there is a teleconnection in summer, indicating that the 48 teleconnection is the relationship between the plateau monsoon, the East Asian monsoon and the 49 South Asian monsoon. In the past 47 years, the Tibetan plateau has shown a significant warming 50 trend and increased precipitation. (Li et al., 2010) The thermal effects of the Tibetan Plateau not 51 only have an important impact on the Asian monsoon and precipitation variability, but also affect 52 the atmospheric circulation and climate in North America and Europe and the South Indian Ocean 53 https://doi.org/10.5194/hess-2020-360 Preprint. Discussion started: 25 August 2020 c Author(s) 2020. CC BY 4.0 License.

54
The various thermal and dynamic effects of the Tibetan Plateau on the atmosphere affect the 55 free atmosphere through the atmospheric boundary layer. Therefore, it is particularly important to 56 analyze the micro-meteorological characteristics of the atmospheric boundary layer of the 57 Qinghai-Tibet Plateau, especially the near-surface layer. (Li et al., 2000). Affected by the unique 58 underlying surface conditions of the Tibetan Plateau, local heating shows interannual and 59 interdecadal variability (Zhou et al., 2009). Land-atmosphere interaction refers to a series of 60 complex processes such as thermodynamics, dynamic, hydrology, biophysical and biochemical 61 processes that occur on land surface, and the interaction process between these processes and the 62 atmosphere (Su et al., 1999). Different underlying surfaces have different diversity, complex 63 composition and uneven distribution. They also make the land surface composed of them diverse 64 and complex. As the main input of atmospheric energy, the surface greatly affects the various 65 interactions between the ground and the atmosphere, and even plays a key role in local areas or 66 specific time (Guan et al, 2009). For this reason, the study of the land-atmosphere interaction on 67 the Tibetan Plateau has become one of the research hotspots in the past 30 years, and has received 68 more and more attention on the whole world.

69
The climate system is sensitive to anomalous changes in land surface conditions. The surface 70 characteristic parameters (dynamic roughness, thermodynamic roughness, etc.) play an important 71 role in the land surface process and are important factors in causing climate change (Jia et al.,  intensity and interaction between the near surface airflow and the underlying surface to some 82 extent. (Liu et al., 2007;Irannejad et al, 1998;Shao et al, 2000;Zhang et al., 2003). Zhou et al. 83 (2012) demonstrated that simulated sensible heat flux compared with measurement was 84 significantly improved by using a time-dependent z0m. Therefore, the primary objective of this 85 study is to calculate the surface roughness and its variation characteristics so that furthermore 86 understanding of land-atmosphere interactions on the central of the Tibetan Plateau.

87
The Nagqu area is located in the hinterland of the Tibetan Plateau and is located in the the 88 central of the Tibetan Plateau. It is the source of the Yangtze River, the Nujiang River, the Lhasa 89 River and the Yigong River. The whole terrain is inclined to the west and the east is low, with an 90 average elevation of over 4,500 meters. The altitude is high, the heat is insufficient, the climate is 91 cold and dry, and the oxygen content is only half of the sea level. The Tibetan Plateau is one of the 92 worst climatic conditions in Tibet. It is a typical sub-cold climate zone. Alpine and hypoxia, dry 93 climate, large temperature difference between day and night, windy weather. (Li et al., 2004) 94 However, there is a vast natural grassland here, so a complete mesoscale observation network 95 centered on the Nagqu climate observation and research station has been established, and a large 96 amount of valuable observation data has been obtained for more accurate description to provide 97 sufficient evidence for Plateau land-atmosphere interactions and atmospheric boundary layer 98 structures. The underlying surface of the Nagqu area is a vast highland plain, and the vegetation is 99 alpine grassland. However, the underlying surface has different degrees of ups and downs and has 100 certain complexity, which brings certain difficulties for the meticulous profound study of the 101 land-atmosphere interaction on the Tibetan Plateau.

102
In this study, the satellite data is obtained by MODIS, and the normalized difference 103 vegetation Index(NDVI) is studied through the Nagqu area to obtain the dynamic surface  land-atmosphere interaction in the plateau region, and to improve the theoretical research level of 111 the near-surface layer in the Tibetan Plateau. In the following section, we describe the case study 112 area, the MODIS remote sensing data, the ground observations, and the land cover map used to 113 drive the revised Massman model (Massman et al, 1997(Massman et al, , 1999. In Section 3, we present the

129
The BJ station is located at 31.37°N, 91.90°E, and the altitude is 4509m a.s.l. The BJ 130 observation point is located in the seasonal frozen soil area, and the vegetation is alpine grassland.

131
The site measurement equipment includes ultrasonic anemometer (CAST3), CO2/H2O infrared 132 open circuit analyzer (LI 7500) and automatic meteorological observation system. (Ma et al., 2006)  172  was selected and the dynamic variation of the surface roughness was obtained.

173
According to the Monin-Obukhov similarity theory (Monin et al., 1954), the wind profile 174 formula with the stratification stability correction function (Panosky et al, 1984 Where: Z0m is the dynamic surface roughness length; z is the observed height; d is zero 179 plane displacement, taken as 0.03m (Yang et al., 2014), calculated from the average vegetation 180 height 0.045m; U is the average wind speed; k is the Karman constant, taken as 0.35;

188
For fully covered uniform canopy， Brutsaert suggest that z 0m =0.13hv (Brutsaert, 1982). For 189 the canopy with proportional coverage (partial coverage), according to the research of Raupach 190 (Raupach, 1994), z 0m/hv varies with the leaf area index I. However, Pierce et al. (Pierce et al. 191 1992) pointed out that for all kinds of biological groups, leaf area index can be obtained from 192 NDVI, and the density grade of vegetation can be related to NDVI. Asrar et al (Asrar et al., 1992) 193 pointed out that there was mutual relationship between LAI, NDVI and ground cover through the 194 study of physical model. Moran's study (Moran, et al., 1994) gives another way, using a function 195 of the relationship between NDVI and Z0m in the growing season of alfalfa, Where 1 and 2 are empirical constants.

198
Considering that the main underlying surface of the study area is grassland, this study selects 199 the Massman model (Massman et al, 1997(Massman et al, , 1999      It can also be seen from Fig. 3, Fig. 4, Fig. 5 that Z0m changes with spatial and temporal 269 scale. Z0m shows different trends on different underlying surfaces. It is worth noting that in 270 November 2008, the Z0m in the Nagqu area was small overall, generally as low as 1 cm.

271
According to historical data, it is known that there is a large-scale snowfall process in the Naqu 272 area at this time. The snowfall of the meadow causes the underlying surface of the meadow to be 273 homogeneous and flat, and after the snowfall falls, it is easy to form a block, scattered and 274 discontinuous underlying surface. And it can be obtained later that the surface roughness of the 275 area with ice and snow as the underlying surface is not more than 1 cm, which is consistent with 276 the historical weather process. So think that snowfall caused the Z0m in November to be very 277 small. And from November to December, Z0m showed a growing trend, which may be due to 278 temperature, soil unfrozen or other reasons, resulting in the melting of snow, and then the surface 279 roughness showed a growing trend (Zhou, 2017

342
The Z0m of the bare soil is at the lowest point of these underlying surface Z0m, and the Z0m of 343 the alpine meadow is relatively stable and less affected by the outside. The fourth category is ice 344 and snow, including ice sheets and snow cover, water bodies two kinds of underlying surfaces, 345 accounting for 3.7% of the area. The Z0m of these three underlying surfaces presents another 346 phenomenon. The variation range of the whole year is relatively small, and the Z0m of these 347 underlying surfaces is also small. It is more than 1cm in mid-June, and other times Z0m is less 348 than 1cm. Figure 8