Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Chine → 2017 → Parameterizations and Modeling of Land-Surface Process over Hinterland of Taklimakan Desert

Université du Xinjiang (2017)

Parameterizations and Modeling of Land-Surface Process over Hinterland of Taklimakan Desert

李火青;

Titre : Parameterizations and Modeling of Land-Surface Process over Hinterland of Taklimakan Desert

Auteur : 李火青;

Grade : Master’s Theses 2017

Université : Université du Xinjiang

Résumé
Land surface process is closely related to human production and life,and it is one of the most important processes in the interaction of energy system in the earth system.At present,the mechanism of land surface process is not enough perfect,and there is a lot of advancement for improvement,land model as an effective tool to research and describing material and energy interactions between land and atmosphere,which helping us understand the mechanisms of the Earth system and global climate change.Furthermore,key parameters and parameterization schemes of land surface serve as the main part of the land model.In this research,Taklimakan Desert hinterland was choosed as an ideal research target area due to it’s characteristic geographical and climate environment.Fisrt of all,we introduced the mechanism of energy balance,water cycle and momentum interaction of CoLM2014,and then identified four key parameters and two parameterization schemes : dynamic roughness,thermal roughness,emissivity,abledo,soil heat conductivity and soil heat volumetric capacity.Next,we preprocessed the data the observation data from Tazhong observation station,after analysis and calculation,the optimal parameters and parameterization schemes were acquareed,in the meantime,we compared and analyzed of the difference of them with Co LM2014 default.At last,we revised parameters and parameterization schemes above mentioned in CoLM2014 and renamed it as Co LM2014_Revised.And then we used same forcing data of Tazhong station to drive Co LM2014 and CoLM2014_Revised respectively,the results of soil temperature,moisture,soil heat flux,net radiation and sensible heat flux were shoosed for comparison and analysis with observation data.The main conclusions as follows :(1)The dynamic roughness z0m=2.991×10-3 m was obtained by the method of Monin-Obukhov similarity theory with the atmospheric stability correction function.(2)The thermodynamical roughness scheme z0 h of Y07 is more suitable for arid area,through drawing out curves of the thermodynamical modeling of default scheme Z98 and Y07 roughness,we found that ln(z0h)value hourly diurnal of Y07 scheme can reflect the z0 h change over time.(3)Based on the surface spectral observations data of the Taklimakan Desert hinterland,and combined with the satellite MODIS emissivity product,we developed an empirical regression equation to estimate the surface broadband emissivity(ε)for the spectral domains 8–13.5 μm by the MODIS data.At last,we used the optimal regression equation to estimate the surface broadband emissivity of Xinjiang,and ε=0.905 in Tazhong area by calculation.(4)Based on the radiation data of the observation station over Tazhong station,the daily mean value of albedo during 2014 was calculated,according to the principle of albedo value variation depend on time factor,we found it is positive related to the sun zenith,so the albedo parameterization equation include time(Day of Year,DOY)factor was obtained : ?=0.273(10)0.022sin ?2 ?(7)d(10)82(8)365.25 ?? ?.(5)According to the heat conduction equation and the observed data over Tazhong station,the soil thermal conductivity at the depth of 2.5,8,20 cm was calculated,and the above three layers soil thermal conductivity was interpolated to 10 layers default by the nonlinear interpolation algorithm with the soil depth.(6)It is recognized that the heat volume capacity cv = 1.599 J·cm-3·K-1 is the most suitable value for the shallow soil in the Tazhong station.We compared the difference of parameters and parameterization schemes with CoLM2014 default.Through modeling and analysis,the fact was proved that Co LM2014 has certain modeling capability in the hinterland of the Taklamakan Desert by comparing the modeling results and observation data.The results of daily and hourly modeling can reflect the changing trendence of surface temperature,soil temperature,soil heat flux and net radiation.Especially for 10 cm and 40 cm soil temperature modeling accuracy is higher,R2 were 0.986 and 0.993 respectively.But the results of heat flux,surface temperature,soil heat flux of modeling accuracy is poor.At noon,the sensible heat flux modeling results are lower than observations ;surface temperature at the beginning and end of the year there are frequent overestimation phenomenon ;soil heat flux in the spring modeling of the overall high,lower winter.The geographical environment of the Taklimakan Desert hinterland is simple,there are no vegetation cover and tiny precipitation,substance and energy cycle without vegetation and other biological participation,so the modeling results is not ideal enough Co LM2014.The RMSE of modeling results of the CoLM2014_Revised are lower than that of the original Co LM2014.The NSE of CoLM_Revised also has been improved obviously,especially the soil heat flux,the NSE value is not acceptable before the correction parameter,but the revised parameters have been significantly improved,within the acceptable range.On the whole,the modeling results of the modified parameters are generally improved,which shows that the revised parameters and parameterization schemes have a positive effect on the modeling results.Finally,we calculated the ratio of energy balance simulation results show that the energy closure rate of the modeling results was improved by 6.1%,so the performance of the CoLM2014_Revised has comprehensive improvement.

Mots clés : Land model; CoLM2014; Parameterization; Modeling; Taklimakan Desert;

Présentation (CNKI)

Page publiée le 31 janvier 2018