ECCE Summer School for Advanced Study in Climate and Environment 2006年7月30-8月12,北京 陆面水文模型发展、参数标定与移植 及其模拟研究 谢正辉，及其研究小组 中国科学院大气物理研究所 http://web.lasg.ac.cn/staff/xie/xie.htm

相关研究小组人员 田向军, 梁妙玲, 张生雷, 袁 飞, 师春香 郑 婧, 宋丽叶 袁 星, 陈 锋 苏凤阁,杨宏伟(USA)

陆面过程 陆面过程是能够影响气候变化的发生在陆地表面的土壤中控制陆地与大气之间动量、热量及水分交换的那些过程；

提 纲 陆面水文模型发展、参数标定与移植 及其耦合、模拟研究； 基于全国50 kmX50 km大尺度陆面水文模型； 提 纲 陆面水文模型发展、参数标定与移植 及其耦合、模拟研究； 基于全国50 kmX50 km大尺度陆面水文模型； 陆面模型的参数标定、移植与模拟； 讨论

陆面过程中地下水位的动态表示及其与气候模式的耦合. Xie Zhenghui, Zeng Qingcun, Dai Yongjiu, and Wang Bin, Numerical simulation of an unsaturated flow equation, Sciences in China(Series D), 4(14),429-436, 1998. Xie Zhenghui, Zeng Qingcun, Dai Yongjiu, An unsaturated soil flow problem and its numerical simulation, Advances in Atmospheric Sciences, 16(2), 183-198,1999 Xie Zhenghui, Liang Xu, Zeng qingcun, A parameterization of groundwater table in a land surfacee model and its applications, Chinese Journal of Atmospheric Sciences, 28(4),331-342, 2004. Liang Xu, Xie Zhenghui, A new parameterization for surface and groundwater interac -tions and its impact on water budgets with the variable infiltration capacity(VIC) land surface model, Journal of Geophysics Research,108(D16), 8613,doi:10.1029/2002-JD003090, 2003. Yang Hongwei, Xie Zhenghui, A new method to dynamically simulate groundwater table in land surface model VIC, Progress in Natural Progress,13(11), 819-825, 2003. Yeh et al 2005 JC. Maxwell et al 2005, JHM. Xie Zhenghui, Xiangjun Tian, Hongwei Yang, A land surface parameterization scheme with a groundwater model for climate models and its applications,2006. Tian xiangjun, Xie Zhenghui, Coupling a Groundwater Component to the NCAR Community Atmosphere Model,2006.

地表地下陆面水文机制 Liang Xu, Xie Zhenghui, 2001, A New Surface Runoff Parameterization with Subgrid -Scale Soil Heterogeneity for Land Surface Models, Advances in Water Resources, 24(9-10), 1173-1193, 2001. Xie Zhenghui, Su Fengge, Liang Xu, Zeng Qingcun, et al,Applications of a surface runoff model with Horton and Dunne runoff for VIC, Advances in Atmospheric Sciences. 20(2), 165-172, 2003. Liang Xu, Xie Zhenghui, Important factors in land-atmosphere interactions: surface runoff generactions and interactions between surface and groundwater, Global Planetary Change, 38,101-114,2003. Tian Xiangjun, Xie Zhenghui, Zhang Shengle, Liang Miaoling, A subsurface ruoff parameterization with water storage and recharge based on the Boussinesq-Storage Equation for a Land Surface Model, Science in China (Series D), 2006.

陆面水文生态模拟 Xie Zhenghui, Liu Qian, Su Fengge, An application of the VIC-3L land surface model with the new surface runoff model in simulating streamflow for the Yellow River basin, IAHS Publiction No.289, 241-248, 2004. 谢正辉，刘谦，袁飞，杨宏伟，基于全国50km×50km网格的大尺度陆面水文模型框架,水利学报，(5),76-82,2004. Yuan Fei, Xie Zhenghui, Liu Qian, Yang Hongwei, Su Fengge,et al, An application of the VIC-3L land surface model and remote sensing data in simulating streamflow for the Hanjiang River Basin, Canadian Journal of Remote Sensing, 30(5), 680-690,2004. Su Fengge, Xie Zhenghui, A model for assessing effects of climate change on runoff in China, Progress in Natural Progress, 13(9), 701-707,2003. 梁妙玲,谢正辉,我国气候对植被分布和净初级生产力影响的数值模拟,气候与环境研究,已接受,2006. Yuan Fei, Xie Zhenghui, Liu Qian, Xia Jun, Simulating Hydrologic Changes with Climate Change Scenarios in the Haihe River Basin, Pedosphere, 15(5): 595-600, 2005.

陆面过程研究前沿问题 水文过程研究需要深入； 生态过程机制（C,N循环）需要发展，植被动态演替； 各种非均匀性问题； 陆面模型的参数标定与移植； 陆面数据同化问题，全球土壤湿度等陆面分量的时空分布； 与区域与全球气候模式的耦合； 各种应用问题； 雪盖、冻土和旱土、大面积水面作用的描述简单，冻土、雪盖占陆面面积都远大于1/4，沙漠区占1/4。

基于全国50 kmX50 km网格 大尺度陆面水文模型

基于全国50 kmX50 km网格 大尺度陆面水文模型 水分收支过程 能量收支过程 陆面过程模式（VIC）

Three-Layer Variable Infiltration Capacity Grid Cell Energy and Moisture Fluxes Grid Cell Vegetation Coverage 2 1 ... 1 N N+1 P Rs RL RL Variable Infiltration Curve Ec S Three-Layer Variable Infiltration Capacity (VIC-3L) Model L i=im[1-(1-A)1/b] im Et i0+P W R P Infiltration Capacity Eb i0 W0 Canopy R As 1 Fraction of Area i i Layer 1 Qb Qd Baseflow Curve Layer 2 Qd Dm Layer 3 Baseflow,B Ds/Ws=1 Ds/Ws<1 Qb DsDm WsW3c W3c Layer 3 Soil Moisture,W3

水分收支过程 植被蒸散、裸土蒸发、土壤水传输、排水和径流决定了陆面过程中的水分收支，也是VIC中所考虑的主要水文过程。 蒸散发（evaporation and transpiration） 冠层截流（canopy interception） 土壤水模型（soil hydrological model） 径流和排水（runoff and drainage）

蒸 散 发 陆面过程模式VIC中的蒸发 冠层湿部蒸发Ew（wet canopy evaporation） 冠层蒸腾Etr（dry canopy transpiration） 裸土蒸发Eg（bare soil surface evaporation）

冠层水量平衡 冠层持水量Mc的平衡方程可由下式表达： P-降水率； Ew -土壤湿部蒸发； Dc-大于叶片最大持水量而滴落到地面的部分。

土壤水模型 用一维Richard’s方程来描述土层间的传导和扩散过程: 各土层的控制方程为： E R P K2 K1 D2 D1 Qb z=0 土壤水模型 z=-z1 z=-z2 用一维Richard’s方程来描述土层间的传导和扩散过程: 各土层的控制方程为： z=-z3

径流和排水 R Infiltration Capacity i0+P P i0 As W0 1 i=im[1-(1-A)1/b] Variable Infiltration Curve i=im[1-(1-A)1/b] Baseflow Curve im Dm i0+P W R P Infiltration Capacity i0 Baseflow,B Ds/Ws=1 Ds/Ws<1 W0 DsDm WsW3c W3c As 1 Fraction of Area Layer 3 Soil Moisture,W3

能量平衡方程 Rn - net radiation; H - the sensible heat flux; E -the latent heat flux; G - the ground heat flux.

VIC 模型结构的简单介绍 土壤、植被参数 源程序及控制文件 Forcing data Flux data

运行VIC模式所需的子目录 ../SOURCE/ 存放源程序及global ../PARAMETER/ 存放土壤及植被参数 ../FORCING/ 存放 forcing data 数据 ../RESULTS/ 存放输出的结果 控制文件

../Source/global 控制文件 在 global 控制文件中包含了运行 VIC 模式所需的三个参数文件： 1、植被参数文件 2、土壤参数文件 3、植被参数库文件 其中，土壤和植被参数文件中存放研究区域内，每个网格中所包含的土壤、植被的相关的统计数据。植被参数库文件存放各种植被类型的一些固定参数。

../Source/global 控制文件 在 global 控制文件中包含了运行 VIC 模式所需的三个参数文件： 1、植被参数文件 2、土壤参数文件 3、植被参数库文件

统计结果如下：

植被参数文件的结构

../Source/global 控制文件 在 global 控制文件中包含了运行 VIC 模式所需的三个参数文件： 1、植被参数文件 2、土壤参数文件 3、植被参数库文件

统计结果如下：

土壤参数文件的结构

../Source/global 控制文件 在 global 控制文件中包含了运行 VIC 模式所需的三个参数文件： 1、植被参数文件 2、土壤参数文件 3、植被参数库文件

../Parameter/参数文件 将生成的参数文件放在 ../Parameter/ 目录下，然后在global控制文件中指定它们所在的目录即可： SOIL ../Parameter/土壤参数文件 VEGPARAM ../Parameter/植被参数文件 VEGLIB ../Parameter/植被参数库文件

../Forcing/forcing data数据

Regional Parameter Estimation of the VIC Land Surface Model: Methodology and Application to River Basins in China Zhenghui Xie, Fei Yuan Institute of Atmospheric Physics Chinese Academy of Sciences, Beijing 100029, China Qingyun Duan University of California/Lawrence Livermore National Laboratory, Livermore, CA 94550, USA Jing Zheng, Miaoling Liang, Feng Chen accepted by Journal of Hydrometeorology   

Model and Data Calibration and transfer Simulation Conclusions

Model and Data Liang et al. (1994) developed the VIC-2L model which includes two different time scales (fast and slow) for runoff to capture the dynamics of runoff generation. To better represent quick bare soil evaporation following small summer rainfall events, a thin soil layer is included in VIC-2L, and VIC-2L becomes VIC-3L. Liang and Xie (2001) developed a new parameterization to represent the Horton runoff mechanism in VIC-3L and combined it effectively with the original representation of the Dunne runoff mechanism(Xie et al., 2003).

Runoff and drainage R=R1(y)+R2(y) i=i m[1-(1-A)1/b] f = f m[1-(1-C)]1/B] i m R2 f m Potential infiltration rate [L/T] Soil moisture capacity [L] y W P i 0 R1 R2 /t wp Wt W/t A C As 1 (a) 1 Fraction of studied area (b) Fraction of the area (1-As)

Saturation excess runoff R1(y) where i0 -- the point soil moisture capacity im -- maximum soil moisture capacity b -- shape parameter(soil moisture capacity) P --precipitation

Infiltration excess runoff R2(y) where fmm -- the average potential infiltration rate fm – the maximum potential infiltration rate B -- shape parameter(potential infiltration rate) P --precipitation ∆t--time step

NSRM 计算示意图 yes no R2 R1 R2 R1 P P As As W0 W0 1 1 Begin Precitation P P+i0<im i i0 +P i im im R2 i0 +P i0 R1 W Y R2 P Ssoil moisture Capacity Ssoil moisture Capacity R1 W Y i0 P W0 W0 As As 1 1 Fraction of Area Fraction of Area Solve Y Infiltration excess runoff R1 Saturation excess runoff R2 No Last time step ? Yes Stop

How to estimate fm From We get tf, then fmm

Infiltration Rate (mm/h) Example: Philip Infiltration Curve f(t) f0 Infiltration Rate (mm/h) W0 t tf where f(t) ---- the infiltration capacity[L/T] Kp---- the final capacity[L/T] Sp---- an empirical constant Time (hour)

Coupling of VIC and NSRM Precipitation Upper layer soil moisture NSRM Surface runoff VIC … Next step

Data and model parameters Vegetation data Soil data Forcing data 50 × 50 km2 resolution

海河流域 黄河流域 淮河流域 长江流域

Vegetation related parameters

Soil classification is based on global 5-min soil data provided by the NOAA hydrology office Soil parameters are derived based on the work of Cosby et al. (1993) and Rawls et al. (1993).

Model parameters of VIC-3L to be calibrated Three depths of three soil layers The exponent of the VIC-3L soil moisture capacity curve B The parameters in the ARNO subsurface flow parameterization

Forcing data are based on 740 meteorological stations in China, which contain 11 years of daily precipitation and air temperature data from 1980 to 1990. Such station information is mapped to the resolution of 50 × 50 km2 grids by combining interpolation methods

Calibration and transfer Classification of climate zones－ Köppen Classification Climatic characteristics for the transfer of calibrated parameters under the premise that hydrological processes and the parameters used to describe them are similar within than between different climate zones Method

Grouping of Köppen climate zones into parameter transfer zones

The climate zones of China according to Köppen classification   The climate zones of China according to Köppen classification

Locations of the selected basins in China for calibration and verifications

Schematic representation of the parameter regionalization scheme

Selected river basins

Parameter calibration Calibration was performed and focused on matching the total annual flow volume and the shape of the monthly hydrograph according to the following procedures: Set the estimated values for the depths of the three soil layers, with deeper depths for arid and semi-arid regions and lower depths for humid regions; Calibrate the ARNO model parameters to fit the low flow; Calibrate the infiltration parameter to match the observed flow peaks, with a higher value to increase the peak and a lower value to lower the peak; Make a fine adjustment on these parameters to get best simulation results.

Parameter transfer scheme Parameters to be transferred Transfer scheme Transferred from the primary to the secondary catchments based on climate zone

Model parameters of VIC-3L to be transferred Three depths of three soil layers The exponent of the VIC-3L soil moisture capacity curve B The parameters in the ARNO subsurface flow parameterization

Parameter transfer (1) Those for two catchments in the Yellow River Basin are calibrated, the parameters for the two catchments are averaged respectively as the corresponding parameters for the zone of continental climate with cool summer. (2) Those for two catchments in the Haihe River Basin are calibrated, and the parameters for the two catchments are averaged respectively as the corresponding parameters for the zone of continental climate with hot summer.

(3) Those for one catchment in the Heihe River Basin are calibrated, and the parameters for the catchment are set to those corresponding parameters for the zone of continental climate with short cool summer. (4) Most of area in the Huaihe River Basin and the Yangtze River Basin belongs to the zone of rainy, mid latitude climate. The parameters for two catchments in the Huaihe River Basin are calibrated, and those for the two catchments are averaged respectively as the corresponding parameters for the zone of rainy and mid latitude climate located in the Huaihe River Basin. Those for two catchments in the Yangtze River Basin are calibrated, and the parameters for the two catchments are averaged respectively as the corresponding parameters for the zone of rainy and mid latitude climate located in the Yangtze River Basin.

Parameters for the rainy and mid latitude climate zone north of the Huaihe River Basin and the Yangtze River Basin are set to that for the Huaihe River Basin; parameter values for the climate zone south of these two river basins are equivalent to that for the Yangtze River Basin. (5) The zone of tropical climate has similar climatic characteristics as those in rainy and mid latitude climate zone. Therefore, the parameters for the zone of tropical climate are set to be the corresponding parameters for the Yangtze River Basin. (6) Since streamflow data for the zone of dry and cold climate is not available, default values of B, D1, D2, Dm, Ds and Ws for the area are set to be 0.3, 0.1, 0.5, 2.0, 0.02, 8.0, and 0.8 respectively.

Simulation Primary Catchments The VIC model also provides a default parameter set, namely the parameter set for base case. Comparisons were made between the results for the base case and calibration.

Mean monthly hydrographs of observed and simulated flow for the primary basins　

Monthly hydrographs of observed and simulated flow for the primary basin　

Mean monthly hydrographs of observed and simulated flow for the primary basins　

Primary Catchments The model performance was considerably better for the calibrated parameters than those without calibration. In general, calibration improve the results in all instances compared those with no calibration.

Secondary Catchments The parameters were transferred to secondary catchments and runoff simulation with the transferred parameters was performed. Runoff simulation with the recalibrated parameters was performed. Results were compared for the base case, the transferred case and the recalibration case.

Mean monthly hydrographs of observed and simulated flow for the secondary basins　

Monthly hydrographs of observed and simulated flow for the secondary basins　

Secondary Catchments The parameter transfer scheme improved the streamflow simulation. Subsequent recalibration of all basins further enhanced the modeling performance.

Calibration and parameter transfer statistics

Summary &Conclusions A parameter estimation is given to simulate streamflow for river basins in China, which is represented by 4355 cells with a resolution of 50 × 50 km2 for each cell. The land area in China was grouped by climate zone, and model parameters were transferred within zones.

The transferred parameters were used to simulate the water balance in river basins in China. The simulated daily runoff of VIC-3L with transferred parameters and un-calibrated parameters was routed to the outlets of the river basins, and compared to the monthly-observed streamflow at the related catchments.

Results show that the model for the transferred parameters can simulate the observations well The parameter transfer scheme improved the streamflow simulation. Subsequent recalibration of all basins further enhanced the modeling performance and the proposed parameter transfer method is promising in estimating the VIC model parameters for data-sparse areas in China.

Discussions 1) what are differences between remote sensed data and model simulated data? 2) why? 3) what is your plan to use the remote sensed data to initialize your model? 4) how can we be coupled? 5)what we do next.

Thank You for your attention! 谢谢各位！ zxie@lasg.iap.ac.cn