Brief Description & Validation Report

Japan

Brief Description & Validation Report March 9, 2020 (ver. 2.1)

Today’s Earth Developing Group

TABLE OF CONTENTS

1. What is “Today’s Earth - Japan”? ...... 1

2. Target of TE-Japan ...... 2

3. Outline of TE-Japan system ...... 2

4. Variable list of TE-Japan ...... 5

5. Validation results ...... 7

i. Snow Amount ...... 7 ii. Soil Moisture...... 9 iii. River Discharge ...... 12 6. Summary of the validation ...... 14

7. Terms of use ...... 16

i. Site Policy ...... 16 ii. User Registration ...... 16 iii. Deletion of User Registration ...... 16 iv. Protection of Personal Information & Handling of Personal Information ...... 16 v. Management of account and password ...... 16 vi. Ownership of Data etc...... 16 vii. Change of the Service ...... 17 viii. Termination of the Service ...... 17 ix. How to cite ...... 17 x. Disclaimer ...... 17 xi. Contact Information ...... 17 8. Member list of TE developing group ...... 18

9. References ...... 19

1. What is “Today’s Earth - Japan”?

“Today’s Earth” (hereafter called “TE”) is JAXA's simulation system of land surface, river discharge and flood area fraction etc. Various products relating to the conditions of land surface and river are obtained as the results of numerical land surface model and river model. “Today’s Earth - Japan” (hereafter called “TE-Japan”) is the regional version of TE, which enables us to see the detailed land surface state of Japan (see Figure 1 for the comparison with TE-Global). Since the target of the TE-Japan is to produce and evaluate long-term high resolution land water cycle data set to calculate risk indices of water hazards, we use MSM- GPV (JMA’s Meso-Scale Model-Grid Point Value) atmospheric analysis data as forcing data of baseline experiment. Furthermore, we are going to utilize satellite data sets developed by JAXA/EORC to replace some input parameters of MSM ver. Rainfall from the Global Satellite Mapping of Precipitation (GSMaP) and solar radiation from Himawari satellite provided by the JAXA Himawari Monitor system are going to be used for satellite version in near future. The TE developer group consists of the researchers in Japan Aerospace Exploration Agency (JAXA) Earth Observation Research Center (EORC), The University of Tokyo (UT), and Remote Sensing Technology Center of Japan (RESTEC). Detailed member list is attached at the end of this document.

Figure 1. Comparison between simulated images of river discharge by TE-Global (0.25-deg lat/lon) and TE-Japan (1/60-deg lat/lon) 1

2. Target of TE-Japan

TE-Japan system aims to produce and evaluate global long-term land water cycle dataset that is helpful for calculating risk indices of water hazards, particularly floods. This activity is designed to broadly contribute the society as a part of the climate services.

3. Outline of TE-Japan system

The TE system consists of land surface model MATSIRO[1] (Minimal Advanced Treatments of Surface Interaction and Runoff) version5[2] and river routing model CaMa-Flood[3] (Catchment-based Macro-scale Floodplain). By giving forcing of surface meteorological parameters, MATSIRO simulates the water and energy interactions between a land surface with a vegetation canopy and atmosphere. The surface runoff and baseflow were calculated independently using Horton flow and the advanced application in TOPMODEL[4], respectively. Based on the calculated runoff amount, CaMa-Flood enables hydrodynamic simulation with floodplain. The model solves the local inertial equation[5], considering a rectangular river channel and trapezium flood plain storage, and represents flood plain dynamics assuming that the elevation profile of the floodplain monotonically increases in each pixel. Severity index of each variables in the form of return period or Mahalanobis distance are also visualized. Table 1 and 2 describe the basic information of TE-Japan. Figure 2 shows the schematic procedure of TE-Japan system.

Table 1. TE-Japan components 1/60° Land Surface Model Horizontal south west lat. = 24°

[1][2] resolution south west lon. = 123° MATSIRO [Nx, Ny]=[1500, 1320] (Minimal Advanced Treatments of Surface Interaction and Runoff) Temporal output every 1 hours starting from resolution UTC+0 each day. 1/60° River Routing Model Horizontal south west lat. = 24° resolution south west lon. = 123° [3] CaMa-Flood [Nx, Ny]=[1500, 1320] (Catchment-based Macro-scale Floodplain) Temporal output every 1 hours starting from resolution UTC+0 each day.

Table 2. Version information of TE-Japan Name MSM-GPV ver.

Period 2007-present

Forcing rainfall M Data*1 snowfall M

eastward wind M

northward wind M

surface air temperature M

specific humidity M

surface shortwave radiation M (downward)

surface longwave radiation M (downward)

surface air pressure M

Data distribution Latency*2 Realtime

1. M: MSM-GPV[6] 2. MSM-GPV forecast data is used to cover the data latency. To distinguish the data using forecast, users are requested to check “Initial_date” described in global_attributes of netCDF data(If the difference between “initial_date” and the date of the file name is within 3 hours, the data is analysis value.). The item of initial_date is included in the data after March 1, 2020. Anything before that will be the analysis value.

Figure 2. Schematic figure of TE-Japan system.

4. Variable list of TE-Japan

All variables distributed from TE-Japan are summarized in Table 3. The column of Fig describes whether figures are available or not through the monitor page.

Table 3. Hydrological variables distributed in the TE-Japan Unit Model/Category Variable Name Item Name Fig (netCDF) rainfall GPRCT kg/m2/s ○ snowfall GSNWL kg/m2/s ○ wind speed GDU m/s - surface air temperature GDT K ○ Forcing specific humidity GDQ kg/kg 〇 surface shortwave radiation (downward) SSRD W/m2 ○ surface longwave radiation (downward) SLRD W/m2 - surface air pressure GDPS hPa - soil moisture (at each level) [Z1-Z6]*1 GLW m/m ○ Water soil moisture (total volume) GLWtot kg/m2 ○ balance (State) canopy water GLWC m - snow amount GLSNW kg/m2 ○ snow melt SNMLT kg/m2/s - snow freeze SNFRZ kg/m2/s - snow sublimation SNSUB kg/m2/s - ice melt ICEMLT kg/m2/s - ice sublimation ICESUB kg/m2/s - snow & ice sublimation SSUB kg/m2/s - 2 Water transpiration ETFLX kg/m /s ○ balance canopy evaporation EIFLX kg/m2/s ○ (Flux) canopy sublimation EISUB kg/m2/s - soil evaporation EBFLX kg/m2/s ○ soil sublimation EBSUB kg/m2/s - MATSIRO total runoff (total) [W1-W2]*4 RUNOFF kg/m2/s ○ base runoff RUNOFFB kg/m2/s ○ surface runoff SRUNOF kg/m2/s ○ runoff (lake & land) [W1-W2]*4 RUNOFFA kg/m2/s - soil temperature [Z1-Z6]*1 GLG K ○ Heat snow temperature [L1-L3]*2 GLTSN K ○ balance *3 ○ (State) land skin temperature [C1-C2] GLTS K canopy temperature [C1-C2]*3 GLTC K ○ soil heat flux GFLUXS W/m2 ○ snow surface heat flux SNFLXS W/m2 ○ ground heat flux in total GFLXTL W/m2 - Heat surface shortwave radiation (upward) SSRU W/m2 - balance surface longwave radiation (upward) SLRU W/m2 - (Flux) sensible heat flux SENS W/m2 ○ latent heat flux LTNT W/m2 ○ latent heat flux (evaporation) EVAP W/m2 -

river flow [W1-W2]*4 RFLOW m3/s - River river water [W1-W2]*4 GDRIV kg/m2 - river storage [W1-W2]*4 GDRIVL kg/m2 - snow covered fraction SNRAT - ○ albedo ALB - - snow albedo [A1-A3]*5 GLASN - - soil potential [Z1-Z6]*1 GPSI Pa - dust density in snow [L1-L3]*2 CDSTM ppmw - water flux atmosphere to land WA2L m/s - water flux land to river WL2R m/s - Others soil ice (at each level) [Z1-Z6]*1 GLFRS m/m ○ soil ice (total volume) GLFRStot kg/m2 ○ land water WLND m - inland water sinkbudget BUDIND kg/m2/s - distributed water sinkbudget RBUDIND kg/m2/s - ground water input WINPT kg/m2/s - lake sh SHLK cm - lake surface temperature TSIL °C - river discharge RIVOUT m3/s - river water storage RIVSTO m3 - river water depth RIVDPH m ○ river flow velocity RIVVEL m/s - floodplain flow (discharge) FLDOUT m3/s - floodplain water storage FLDSTO m3 - CaMa-Flood floodplain water depth FLDDPH m ○ flood area FLDARE m2 - flood fraction FLDFRC - ○ water surface elevation SFCELV m - total discharge (RIVOUT + FLDOUT) OUTFLW m3/s ○ total storage (RIVSTO + FLDSTO) STORGE m3 - 1. Z1-Z6 represents the soil layers, the depth (m) of which is Z1: 0 - 0.05, Z2: 0.05 - 0.25, Z3: 0.25 - 1, Z4: 1 - 2, Z5: 2 - 4, and Z6: 4 - 14. 2. L1-L3 represents the snow layers. The number of the effective layers and their depth are variable. See Takata et al. (2003) for more details. 3. C1 and C2 represent the outputs for snow-free canopy and snow-covered canopy, respectively. 4. W1 and W2 represent the values regarding water and ice, respectively. 5. A1, A2 and A3 represent the snow albedo of visible, near-infrared and infrared area, respectively.

5. Validation results

In order to verify the performance of the TE-Japan products, we had validations for 3 variables: snow amount and soil moisture from MATSIRO, and river discharge from CaMa- Flood. For each variable, we compared the results of TE-Japan and observation data and made some statistical comparison between them. i. Snow Amount

First, we investigated the general characteristics of the results of snow amount. AMeDAS[7] (Automated Meteorological Data Acquisition System) observation of snow depth by JMA (Japan Meteorological Agency) is utilized to validate the TE simulation. Since TE simulates snow amount as a Snow Water Equivalent (SWE) [kg/m2], direct comparison with the AMeDAS snow depth will not be recommended. Here we used the method introduced by Sturm et al. (2010)[8] to convert snow depth to SWE by calculating snow density as a function of snow depth and its aging time. As shown in Figure 3, we picked up 316 AMeDAS observation sites that have snow depth observation from 2011/1/1 to 2015/12/31 in common. Figure 4 is a scatter density diagram during that period, showing the relation between TE-Japan SWE estimates (y-axis) and in-situ SWE converted from observed snow depth (x-axis). Overall, TE-Japan estimates shows good correlation with the in-situ SWE. Looking at the results by each validation site, we found over/underestimation at some sites, however. Aomori site in Figure 4(d) is one example where there is significant underestimation. This can be partly explained by the over/underestimation and/or misalignment of forcing data itself, since the time series of the SWE and snowfall (model forcing) in Figure 5 are in good correspondence, indicating that TE-Japan has the capacity to estimates seasonal variation of SWE to some extent.

Figure 3. Selected AMeDAS observation sites (316 in total). Locations are shown in red dots.

Figure 4. Scatter diagram for the relation between daily SWE of TE-Japan results and in-situ SWE converted from snow depth observation. Colors in each point show the number of samples included. (a) total result of 316 sites, (b) Hakodate site ([lat, lon] = [45.4150, 141.6783]), (c) Wakkanai site ([lat, lon] = [41.8167, 140.7533]), (d) Aomori site ([lat, lon] = [40.8217, 140.7683]), respectively.

Figure 5. Time series of the daily SWE from 2011/1/1 to 2015/12/31 at each site. See Figure 4 for lat/lon information of each site. In-situ SWE are shown in black line, TE-Japan estimates are in blue line. Light blue histograms are daily snowfall of the model forcing.

ii. Soil Moisture

The validation of soil moisture is performed using in-situ observation data obtained from AsiaFlux Database[9]. 11 out of 37 total sites are located in Japan, and 5 sites have overlapped observation with TE-Japan simulation period. We picked up 3 sites that have observation at the depth of 15cm which is same as the middle point of the 2nd layer (root zone) in TE-Japan. The observation site information is summarized in Table 4. Note that average value of 10cm and 20cm depth observation is used to validate the TE-Japan estimates for the FHK and SMF sites.

Table 4. Information of observation sites for soil moisture validation.

Site name Lat/lon Data & Soil Moisture Obs. Period [depth]

Fuji Hokuroku 35.433/138.750 FxMt 2006- volumetric soil water content Flux Observation 2009 [0,10,20cm] Site (FHK) Seto Mixed 35.250/137.067 FxMt 2002- soil water content Forest Site 2011 [2,5,10,20,50cm] (SMF) Takayama 36.146/137.426 FxMt 1998- soil water content deciduous 2007 [15,40cm] broadleaf forest site (TKY)

Similarly as the results of SWE, scatter density diagram to show the correlation between TE-Japan results and in-situ observation for soil moisture at all observation sites in Table 4 is shown in Figure 6.

Figure 6. Scatter diagram for the relation between daily soil moisture of TE-Japan results and in-situ soil moisture. Colors in each point show the number of samples included. (a) shows total results of 3 points, (b), (c) and (d) are the results at FHK, SMF, TKY site respectively.

At TKY site TE-Japan shows almost 1-to-1 relation to the observation, while there are constant positive biases at FHK and SMF sites. These biases can be attributed to the uncertainty of the soil type classification in boundary condition data. For instance, according to the AsiaFlux database document, FHK site is mainly covered with the coarse-grain volcanic ash, while the corresponding grid in TE-Japan is defined as the clay loam. Improvement of the accuracy of soil type classification is one of our future challenges. Figure 7 shows the time series of estimated and observed soil moisture anomaly. We can see that TE-Japan nicely describes soil moisture variation corresponding to the precipitation even at the sites where there is large overestimation.

Figure 7. Time series of the daily soil moisture anomaly from 2007/1/1 to 2007/12/31 at each site. See Table 4 for lat/lon information of each site. In-situ soil moisture are shown in black dots, TE-Japan estimates are in red line. Light blue histograms and blue steps are daily precipitation of the model forcing and in-situ observation respectively. Note that there is no precipitation observation at TKY site.

iii. River Discharge

In this section, we show the validation results of CaMa-flood from 2010 to 2015. Figure 8 shows the validation results of daily river discharge in each observation point. The value of correlation coefficient, root mean square error, mean absolute error, and bias is calculated between the observational discharge, and the TE-Japan discharge results are the nearest grid-point of the observation site. The observation data is obtained from the Ministry of Land, Infrastructure, Transport and Tourism (MLIT). Both high and low performance can be found in the whole country. For all of the rivers, the value of mean absolute error is small, although the river discharge overall is lower than observation, which can be found by Bias. For most of the rivers in Hokkaido, the north part of East Japan, and most rivers in other regions, the value of the correlation coefficient is high, and the mean absolute error is low.

Figure 8. Spatial distribution of validation results for daily river discharge. The observation from MLIT from 2010 to 2015 were compared with the river discharge from TE-Japan shown as circle. Mean correlation coefficient is 0.548, mean of root mean square error is 69.799, mean of mean absolute error is 37.746, and mean bias is -12.859 of MSM version. 12

Inter-annually and monthly-mean river discharge are investigated for four sites: (a) Teshio River, (b) Mogami River, (c) Hanzo River, and (d) Haze River. These four sites are selected from different regions of Japan, located in various sizes of the catchment (Table 5, 6), of which results are relatively accurate. Figure 9 shows a comparison of the results of TE-Japan (red line and blue zone) and MLIT observation data (black line, grey dash and dot line). Most of the seasonal variability is well produced by TE-Japan, although cold bias exists. Next, the reproducibility of inter-annual variability is examined for the same four sites. Figure 10 is the long-term time series of monthly-mean river discharge. Temporal variation of river discharge at the four sites are well produced compared to the observed discharge.

Table 5. Information of observation sites for river discharge validation.

River name Lat/lon Upstream area Data

Teshio River 44.463/142.365 222.00 km2 2010-2015 Mogami River 38.759/140.064 6270.90 km2 2010-2015 Hanzo River 32.835/130.065 87.00 km2 2010-2015 Haze River 34.651/136.425 25.00 km2 2010-2015

Figure 9. The time series of inter-annual (2010-2015) and monthly-mean river discharge at four observation sites, which are from Teshio River, Mogami River, Hanzo River, and Haze River. The black line represents the observation of monthly-mean river discharge from MLIT, and the red line is the result of TE-Japan products of the MSM version.

Figure 10. Time series of monthly-mean river discharge at Haze River, Mogami River, Hanzo River, and Haze River. The red line shows the discharge results of the MSM version of TE-Japan, and the black line indicates the observation data. The results of spatially-averaged precipitation over each basin (Table 5) are shown by blue bars from upside down.

6. Summary of the validation

We showed the validation results of the TE-Japan system, which simulates the regional hydrological cycle on land areas using in-situ observation data. Three variables: snow amount, soil moisture, and river discharge were investigated to verify the performance of the system. The whole

performance is not perfect, but it shows enough accuracy for users to utilize TE-Japan’s data for research or business uses. We are chasing on improving the TE-Japan system by refining the models and boundary conditions, enhancing the resolution of the models, and using advanced forcing data, like stellate-derived observation data effectively. Table 6 shows a summary regarding the performance of the TE-Japan system in terms of three validation variables: SWE, soil moisture, and river discharge.

Table 6. Summary of the validation of TE-Japan Variable Ground Correlation Root mean Bias Mean absolute observation coefficient square error error data Daily average of SWE converted 0.73169 (MSM) 84.731 (MSM) 14.481 (MSM) 55.930 (MSM) SWE (mm) from AMeDAS snow depth obs. (2011 – 2015) Daily average of In-situ 0.53127 (MSM) 14.929 (MSM) 14.385 (MSM) 14.614 (MSM) soil moisture observation data (%vol) by AsiaFlux Daily average of In-situ 0.54783 (MSM) 69.799 (MSM) -12.859 (MSM) 37.746 (MSM) river discharge observation data (m3/s) by MLIT 1. For daily river discharge, correlation coefficient, root mean square error, bias, and mean absolute error are calculated for each observation site and averaged over the whole region.

7. Terms of use

This terms of use is the same as that shown in the Today’s Earth website https://www.eorc.jaxa.jp/water/term.html?1

Today's Earth (hereinafter referred to as "the Service") provides the simulation data of MATSIRO and CaMa-Flood by the Japan Aerospace Exploration Agency (JAXA) Earth Observation Research Center (EORC) and Institute of Industrial Science, The University of Tokyo, in free of charge (only for research, education and public purposes).

Terms of Use of the Service are terms and conditions that users are expected to follow in the use of the Service. Please read the terms and conditions carefully and agree on them before the use of the Service.

i. Site Policy The site policy (http://global.jaxa.jp/policy.html) of the Japan Aerospace Exploration Agency (hereinafter referred to as "JAXA") is applicable to the Service.

ii. User Registration User registration is required before the use of the Service because the Service adopts a user authentication system based on a set of user account and password. The information for the registration is your name, e-mail address, organization, department, country/region, and purpose of data usage. iii. Deletion of User Registration If you want to delete your user registration, please notify the Water Cycle & Water Resource Management Research Group described in the follow links. JAXA reserves the right to delete your user registration if you have not complied with the Terms of Use or are an inappropriate user of the Service. iv. Protection of Personal Information & Handling of Personal Information JAXA handles the personal information such as names and e-mail addresses in accordance with JAXA's rules on protection of personal information. For details, please refer to "Privacy Policy" in the JAXA's site (http://global.jaxa.jp/policy.html).

JAXA doesn't use registered personal information except for the following purposes: ➢ Statistic and analysis of data use ➢ Questionnaire surveys to users for improvement of the Service ➢ Response to inquiries from users In addition, JAXA employs other companies to perform functions on our behalf. The functions include the system management, user management, and the >Water Cycle & Water Resource Management Research Group operation. They may access to personal information to perform the functions, but may not use it for other purposes. v. Management of account and password User account and password are managed and used under the responsibility of user. JAXA is not responsible to you for any loss or damage or due to any other cause beyond the control of JAXA that may be caused by misuse of user account and password by another person. vi. Ownership of Data etc. The copyrights of the standard products and other materials provided in the Service are the property of JAXA. Please use them in compliance with the conditions stipulated in "Scope and

Conditions for Use of the Contents of the Site" in the JAXA's site (http://global.jaxa.jp/policy.html).

vii. Change of the Service JAXA may change the Term of Use, service contents and conditions of the Service. JAXA shall notify users of the change on "Information" of the Service. Please confirm the change when you use this service for the first time after the change. viii. Termination of the Service JAXA reserves the right to terminate the Service by JAXA's own reasons. In that case, JAXA shall notify users of the termination beforehand by proper means.

ix. How to cite Please specify the following sentence when you publish a thesis, a report, and so on by using the research products and images supplied by the Service.

"'Research product of Yesterday's Earth at EORC that was used in this paper was supplied by Japan Aerospace Exploration Agency and Institute of Industrial Science, The University of Tokyo."

The Water Cycle & Water Resource Management Research Group is collecting the related literature. It would be very appreciated if you could send a reprint or a copy of your research outcome to the Water Cycle & Water Resource Management Research Group.

x. Disclaimer Please be aware in advance that: Although JAXA has taken every care to manage the Service, JAXA assumes no responsibility regarding the safety of the contents of the Service or the reliability of information provided on the Service. JAXA is not responsible to you for any damage that may be caused by the use of the Service and/or the information on the Service. JAXA may change or delete the information on the Service, or may suspend or terminate the operation of the Service. JAXA is not responsible to you for any inconvenience that may be caused by such changes or deletion of the information, or by such suspension or termination of the operation of the Service.

xi. Contact Information If there are any concerns or questions about the Service, please contact us at the following links, https://www.eorc.jaxa.jp/water/contact_us.html or send an email to

8. Member list of TE developing group

Latest member list is at https://www.eorc.jaxa.jp/water/contact_us.html

NAME ORGANIZATION ROLE DESCRIPTION Current Members ➢ Overall (management, collaboration with other “Water R. Oki JAXA*1 Cycle and Water Resource Management” topics) ➢ Operation of the TE systems M. Kachi JAXA ➢ Validation of land variables (snow depth, soil moisture) ➢ Collaboration with other EORC research groups H. Fujii JAXA ➢ Operation of the TE systems ➢ Operation of the TE systems K. Yamamoto JAXA ➢ Validation of river variables (runoff, flooding area) T. Oki UT*2, JAXA ➢ Advisory (science & outcomes) K. Yoshimura UT ➢ Overall (Principal investigator at UT) H. Kim UT ➢ Advisory (science & outcomes) D. Yamazaki UT ➢ Improvement of CaMa-Flood model ➢ Development of Japan 1km-resolution TE K. Hibino UT ➢ Further improvement of TE ➢ Collaboration with other EORC research groups ➢ Development of Japan 1km-resolution TE W. Ma UT ➢ Further improvement of TE ➢ Collaboration with other EORC research groups Y. Ishitsuka UMass Amherst*3 ➢ Further improvement of TE A. Takeshima UT ➢ Further improvement of TE R. Arai UT ➢ Further improvement of TE T. Higashiuwatoko RESTEC*4 ➢ Operation of the TE systems T. Andoh RESTEC ➢ Operation of the TE systems R. Kakuda RESTEC ➢ Operation of the TE systems Previous Members ➢ Operation of the TE systems T. Nomaki RESTEC ➢ Validation of land variables (snow depth, soil moisture) ➢ Operation of the TE systems N. Kawamoto RESTEC ➢ Validation of land variables (snow depth, soil moisture) T. Itaya UT ➢ Development of downscaling method ➢ Operation of the TE systems S. Urita RESTEC ➢ Validation of land variables (snow depth, soil moisture) M. Hatono UT ➢ Validation of river variables (discharge, flooding area) Y. Yabu UT ➢ Development of Japan 1km-resolution TE 1. JAXA: Japan Aerospace Exploration Agency 2. UT: University of Tokyo 3. UMass Amherst: University of Massachusetts (Amherst) 4. RESTEC: Remote Sensing Technology Center of Japan 18

9. References

1. Takata, K., S. Emori, and T. Watanabe, 2003: Development of the Minimal Advanced Treatments of Surface Interaction and RunOff (MATSIRO), Global and Planetary Change, 38, 209-222. 2. Nitta, T, K. Yoshimura, K. Takata, R. O’ishi, T. Sueyoshi, S. Kanae, T. Oki, A. Abe-Ouchi, and G. E. Liston, 2014: Representing variability in subgrid snow cover and snow depth in a global land model: Offline validation, J.Clim., 27, 3318–3330. 3. Yamazaki, D., S. Kanae, H. Kim, and T. Oki (2011), A physically based description of floodplain inundation dynamics in a global river routing model, Water Resour. Res., 47, W04501, doi:10.1029/2010WR009726. 4. Beven, K.J., M.J. Kirkby, N. Schofield, A.F. Tagg, 1984: Testing a physically-based flood forecasting model (TOPMODEL) for three U.K. catchments, J. Hydrol, Volume 69, Issues 1–4, Pages 119-143, ISSN 0022-1694, https://doi.org/10.1016/0022-1694(84)90159-8. 5. Bates, P.D., M.S. Horritt, and T.J. Fewtrell, 2010: A simple inertial formulation of the shallow water equations for efficient two‐dimensional flood inundation modelling, J. Hydrol., 387, 33– 45, doi:10.1016/j.jhydrol.2010.03.027. 6. MSM-GPV, http://www.jmbsc.or.jp/jp/online/file/f-online10200.html, 2019. 7. AMeDAS(The Automated Meteorological Data Acquisition System), https://www.jma.go.jp/jma/en/Activities/amedas/amedas.html, 2019. 8. Strum, M., B. Taras, G. E. Liston, C. Derksen, T. Jonas, and J. Lea, 2010: Estimating Snow Water Equivalent Using Snow Depth Data and Climate Classes, Journal of Hydro Meteorology, 11, 1380-1394. 9. AsiaFlux Database, https://db.cger.nies.go.jp/asiafluxdb/?page_id=16, 2019