Hydrological Research Letters, 4, 25–29 (2010) Published online in J-STAGE (www.jstage.jst.go.jp/browse/HRL). DOI: 10.3178/HRL.4.25 Hydrological response to future climate change in the basin, Xieyao Ma1, Takao Yoshikane1, Masayuki Hara1, Yasutaka Wakazuki1, Hiroshi G Takahashi1,2 and Fujio Kimura1,3 1Research Institute for Global Change, Japan Agency for Marine-Earth Science and Technology, Kanagawa, Japan 2Department of Geography, Tokyo Metropolitan University, Tokyo, Japan 3Institute of Life and Environmental Sciences, University of Tsukuba, Ibaraki, Japan

Abstract: requires information on changes in river discharge determined from both observation and simulation. Hydro- To evaluate the impact of climate change on snowfall logical simulations of rivers generally use data obtained in Japan, a hydrological simulation was made in the Agano from ground-based meteorological stations. The distribution River basin by using a regional climate model’s output. A and density of meteorological stations affect the accuracy hindcast experiment was carried out for the two decades of such hydrological modeling, especially for large river from 1980 to 1999. The average correlation coefficient of basins. Japan has approximately 110 manned meteorological 0.79 for the monthly mean discharge in the winter season stations, which measure wind speed, precipitation amount, showed that the interannual variation of the river discharge air temperature, humidity, atmospheric pressure, and other could be reproduced and that the method can be used for variables. Nevertheless, this density does not meet the climate change study. The future hydrological response to recommendations of the World Meteorological Organization global warming in the 2070s was investigated using a (WMO, 1994). In addition, because Japan is a mountainous pseudo-global-warming method. In comparison to data from country, most stations are located in valley bottoms or other the 1990s, the monthly mean discharge for the 2070s was flat areas. projected to increase by approximately 43% in January and In recent years, a downscaling method that links 55% in February, but to decrease by approximately 38% in atmospheric and hydrological models has been developed April and 32% in May. The flood peak in the hydrograph for hydrological simulations. Kite and Haberlandt (1999) was moved forward by approximately one month, changing tested hydrological simulations of the Mackenzie and upper from April in the 1990s to March in the 2070s. Furthermore, Columbia rivers and showed that the coupling of atmospheric the projection for the 10-year average snowfall amount was and hydrological models was useful in understanding the projected to be approximately 49.5% lower in the 2070s macro-scale hydrological cycle. Many other authors have than in the 1990s. also used the downscaling method to investigate changes in river hydrology (e.g. Wood et al., 2004). Fujihara et al. KEYWORDS hydrological simulation; regional climate (2006) examined the influence of global warming on the model; dynamical downscaling; climate water resources of the basin using the change; river discharge Table I. Mean decadal April discharge rates (m3/s) in the INTRODUCTION 1990s for the main rivers flowing into the in the Hokuriku and Tohoku regions and the differences between 1990 and two earlier decades, given as percentages Climate change is having a conspicuous effect on Japan. of 1990s values. In particular, the Japan Meteorological Agency (2002) has Name of 2 reported that snowfall amounts have fallen sharply along Catchment (km ) 1990s Δ (1970s) Δ (1980s) river Japan’s eastern seaboard since the mid-1980s, resulting in a noticeable decrease in river discharge in this region in Kuzuryo*1 Nakatsuno (1,240) 121.1 33% 36% spring. Table I lists decadally averaged April discharge rates Tedori Nakajima (732) 142.1 15% 20% 2 (m3/s) for the main rivers entering the Sea of Japan, with Sho* Daimon (1,120) 64.9 24% 27% differences between the earlier decades and the 1990s given Jintsu Jintsu-ohashi (2,688) 300.1 12% 10% as percentages of the 1990s values. There are no negative Agano Maoroshi (6,997) 674.0 24% 21% Mogami Sagoshi (6,497) 762.0 26% 29% change values in the table, indicating that all of the rivers Omono Tsubakigawa (4,035) 392.7 28% 29% have experienced a decrease in their mean April discharge Yoneshiro Futatsui (3,750) 422.0 11% 16% rate. Indeed, April discharge rates in the 1990s were 10% to 36% smaller than those in the 1970s and 1980s, which *1 1970, 1994 and 1997 data missing; 2 also provides supporting evidence for climate change in this * 1993 data missing; region. 1990s: 10-year monthly mean discharge from 1990 to 1999; To clarify the possible effects of climate change on water 1980s: 10-year monthly mean discharge from 1980 to 1989; 1970s: 10-year monthly mean discharge from 1970 to 1979; resources, quantitative analysis is required. Such analysis Δ (1970s)=100* (1970s–1990s)/1990s; Δ (1980s)=100* (1980s–1990s)/1990s. Correspondence to: Xieyao Ma, Research Institute for Global Change, Japan Agency for Marine-Earth Science and Technology, 3173-25 Showamachi, Kanazawa-ku, Yokohama City, Kanagawa 236-0001, Japan. Received 14 December, 2009 E-mail: [email protected] ©2010, Japan Society of Hydrology and Water Resources. Accepted 3 March, 2010

—25— X. MA ET AL.

downscaling method with bias correction and Tachikawa MODELS AND SETTING et al. (2009) states that the hydrological regime change should be investigated using high resolution general To examine river discharge, we used the WRF regional circulation model (GCM) data without any correction. climate model and the SVAT&HYCY model. Kobayashi et al. (2008) investigated precipitation change over the Shiribetsu River basin (1,640 km2) and Gokase The WRF Model River basin (1,820 km2) and compared simulated The WRF model is a mesoscale numerical weather hydrographs derived from various datasets. Hara et al. (2008) projection system designed to serve both operational reported the impact of global warming on snow depth in forecasting and atmospheric research needs. The model uses Japan. Hara et al.’s (2008) results correlated with observed full-compressible, non-hydrostatic equations. We used the precipitation during the winters of 2005 (high snow cover) Advanced Research WRF (ARW) core version 2.2 and 2006 (low snow cover) and projected that the snow (Skamarock et al., 2005) with a two-way nesting technique depth will decrease by 40% under the warmer conditions and the WRF single-moment 6-class microphysics scheme. currently projected for the 2070s. Wada et al. (2008) The parent domain was set to a wide area, between 28–46°N conducted a flood risk assessment of global warming using and 124–148°E on a 20-km grid. The inner domain was a regional climate model and high resolution GCM over located in a smaller area between 35.5–39.3°N and 135.7– Japan and indicated that the risk will increase in most areas. 141.5°E with a 5-km grid (Figure 2a). As mentioned by Tachikawa et al. (2009), the bias must We conducted two numerical experiments. One was the be removed from the GCM output data in order to do hindcast run (CTL) used to reproduce past hydrological hydrological modeling because the statistical properties of events of the 1980s and 1990s. The other was a pseudo- the GCM output data do not necessarily match up to that global-warming run (PGW) used to project the hydrological of observed data. There is no guarantee that the obtained response in the 2070s. The National Centers for technique, a bias correction technique from the current Environmental Prediction/National Center for Atmospheric climate experimental data and observed data, is applicable Research (NCEP/NCAR) 6-hourly reanalysis data (Kalnay to the future data. In this study, we examined the projections et al., 1996) was used as the lateral boundary condition for of a regional climate model used to investigate long-term the CTL. For the PGW, the lateral boundary condition was hydrological responses using a hydrological model. We used the Weather Research and Forecasting (WRF) model with dynamic downscaling to simulate input variables for the Soil-Vegetation-Atmosphere Transfer and Hydrological Cycle (SVAT&HYCY) model. The study site was the Agano River basin, an area in Japan that receives heavy snowfall, located in the . Model runs over a 20-year period were conducted, and the output hydrographs were compared with observational data. We then examined the effects of global warming on the hydrological processes of the Agano River basin using the pseudo-global-warming method.

STUDY AREA

The Agano River (Figure 1) drains into the Sea of Japan and is the second largest river in Japan with respect to annual discharge (12.9 billion m3, http://www.hrr.mlit.go.jp/agano/). The Agano River drains an area of 7,710 km2 making it the Figure 1. Location of the Agano River basin. eighth largest river basin in Japan. The two major tributaries of the Agano, the Tadami and Aga rivers, are both located in . The majority of the Agano River discharge is caused by the , which drains an area receiving heavy winter snowfall. In its lower reaches, the Agano River flows through the Echigo Plain in Prefecture. The Echigo Plain is a major rice production area, ranking second in Japan in total rice output, and water is an important factor in determining the quantity and quality of the rice. In addition, there are many hydroelectric dams because of the abundant water resources of the basin. Therefore, the projection of future climate change in the Agano River is significant for the region’s economy and society. Figure 2. Domain of the WRF model (a): ① center of the parent domain (20-km grid); ② center of the inner domain (5-km grid). River network of the Agano River basin (b).

—26— HYDROLOGICAL RESPONSE TO FUTURE CLIMATE CHANGE IN THE AGANO RIVER BASIN, JAPAN adopted, following the method of Hara et al. (2008). A (Table S1 and Figure S1 in Supplement), respectively. Figure global warming component was added to the NCEP/NCAR 3 shows the annual runoff from 1980 to 1999. Simulated reanalysis data for the 1990s. Global warming components annual runoff roughly corresponded to observed values were estimated as the monthly average difference between throughout the two decades with the exception of 1996. The the 10-year average of the 21st century projection, based on simulated interannual variation of annual mean runoff scenario A2 of the Special Report on Emissions Scenarios correlated well with the observed data, though the estimated (SRES-A2) (Nozawa et al., 2007), from 2071 to 2080, and precipitation was overestimated compared to some of the the 20th century simulation from 1991 to 2000, from version station data. The time series correlation was significant, with 3.2 of the Model for Interdisciplinary Research on Climate a significant level of 95% (0.62 in correlation coefficient). (MIROC) (medres, approximately 300 km, T42), an Figure 4 shows the monthly mean discharge in winter atmosphere-ocean coupled general circulation model. The from November to May. Although there was no significant PGW method allows for the comparison of climate in the difference between the simulation and observation, it is more present year and that in a PGW year that is similar to the important to examine the correlation rather than the control year in terms of interannual variation while including comparison in absolute values because of human impact. future climatology. Therefore, by the PGW method, we could For example, the water controlled by dams was not evaluate the river discharge under a future climate. considered in this study. The correlation coefficient for the The output variables from the WRF model were (1) monthly mean discharge in average was 0.79 with a range downward longwave flux at the ground surface, (2) from 0.66 in December to 0.84 in February. The correlation downward shortwave flux at the ground surface, (3) surface coefficient was 0.79, 0.75 and 0.83 in March, April and air pressure, (4) water vapor mixing ratio at 2 m height, (5) May, respectively. The average of these time series did not accumulated total grid-scale precipitation, (6) surface air have a significant difference, with a significance level of temperature (SAT), and (7) wind speed at 10 m height. 5%. It suggests that interannual variation is greatly influenced by the climate factor. The calculated monthly The SVAT&HYCY model mean discharge in April in the 1990s decreased 16.6% The SVAT&HYCY model (Ma et al., 2000) is a one- compared with that of the 1980s, which was consistent with dimensional hydrological model that includes a soil- vegetation-atmosphere transfer scheme, runoff formation, and river routing. The model has a number of features applicable to understanding the water-energy cycle in a river basin. In this study, only the forest vegetation type was used. The precipitation was assumed to be snowfall when the SAT was less than 2°C according to the result of Ma et al. (1999). Evapotranspiration, snowmelt, surface runoff, infiltration, and base flow were simulated by the SVAT&HYCY model using the WRF output data listed above. The river network was constructed using the GTOPO30 dataset (http:// edc.usgs.gov/products/elevation/gtopo30/gtopo30.html) with 0.05° resolution (Figure 2b). The estimated basin area was 7,834 km2, which is close to the reported area of 7,710 km2. The velocity of the river flow was set at 0.7 m/s, in consideration of the mountainous topography. The amount of discharge measured at Maoroshi (6,997 Figure 3. Comparison of annual runoff between the 2 km ) until 2003 is available from the Japan River simulation (CTL) and observations (OBS) at Maoroshi. Association. To confirm the performance of the hydrological model, we ran it for a 20-year experiment period from July 1979 to June 1999. The WRF model was run per year, from June 28 of this year to July 1 of the next year. We did not consider the effect of dams in this study because many dams along the river are used to operate with high water levels for hydroelectric power generation. We only focused on the relative evaluation of interannual variation, which is greatly influenced by climate factors in the monthly mean discharge.

RESULTS AND DISCUSSION

To avoid snow pack, the “water year” was set from July of the previous year to June of the current year. The simulated amount of precipitation was compared with 10-station data in the 20-year period. The result showed that the annual Figure 4. Comparison of 10-year average monthly mean mean correlation coefficient and bias were 0.63 and 592 mm discharge between the simulation (CTL) and observations in overestimation (with a standard deviation of 663 mm) (OBS) in the 1980s and 1990s at Maoroshi.

—27— X. MA ET AL. the data and approximately 5% smaller than that shown in Table I. In addition, The Nash-Sutcliffe efficiency coefficient (NSC) (Nash and Sutcliffe, 1970) was adopted to evaluate the reproducibility of the hydrological model. The NSC values for the monthly mean discharge were 0.474 for the 1980s and 0.476 for the 1990s. As Figure 4 demonstrated with the high NSC, the seasonal variation of river runoffs during the period is well reproduced in the hydrological model without dam operations, confirming that the effect of the dam operation is minor during the simulated season. Therefore, it is considered that the model is appropriate for studying climate change impact, especially during the winter season. Figure 5. Same as in Figure 4 but for simulations in the The monthly mean discharge in the 1990s and that under 1990s (CTL) and the 2070s (PGW). the PGW in the 2070s are shown in Figure 5. The figure presents a clear increase in discharge from 340 to 486 m3/s (43%) in January and from 365 to 565 m3/s (55%) in February. However, discharge decreases are shown in both April (from 674 to 417 m3/s, 38%) and May (from 427 to 292 m3/s, 32%). In the 2070s, the flood peak is projected to occur one month earlier, in March rather than April. There was little change between July and November, and in general, the annual mean changes of runoff between the 1990s and 2070s runs were small. This suggests that the total precipitation change between the 1990s and 2070s runs was small. Under global warming, the discharge change is mainly caused by the phase change of precipitation (Hara et al., 2008) and earlier snowmelt due to increased SAT. The characteristics of hydrological response for future climate change obtained in this study are consistent with the result of Tachikawa et al. (2009) for the basin. The 10-year average annual precipitation of 2,243 mm in the 1990s run increased by approximately 80 mm (3.6%) to 2,323 mm in the 2070s run. However, the amount of Figure 6. Distribution of the decrease ratio (percentage) of the 10-year average snowfall amount in the PGW run snowfall showed a marked decrease. The 10-year average compared with the CTL run over the basin. annual amount of snowfall in the 1990s run was 507 mm, which decreased to 251 mm in the 2070s run, representing a basin-wide reduction of 49.5%. Figure 6 shows the the relatively high SAT increase in the 2070s between the distribution of snowfall decrease across the basin. The CMIP3 GCMs. The ratio of snowfall in total precipitation snowfall decreased by more than 20% (ranging from 23% can be greater and the snowmelt season could come later to 95%). Particularly large snowfall reductions of more than when we use the multi-model ensemble mean as a global 80% are shown in the downstream region, mostly because warming component. the PGW model projected no snow there. The changes of all grids over the basin were significant by t-test with the CONCLUSION significance level of 5%. The PGW components by AOGCM that we used in this study are obtained from MIROC (Nozawa et al., 2007). The To understand the impact of climate change on future change of the annual precipitation around Japan is hydrological processes in snowy areas of Japan, a around zero by not only MIROC and but also the ensemble hydrological model was applied to the Agano River basin mean of the CMIP3 GCMs (Chapter 11 of IPCC, 2007). by using dynamic downscaling data. The model performance The difference of the 10-year-averaged global mean SAT of was checked using a 20-year hindcast for the period 1980 CMIP3 GCMs between 1990s and 2070s is 2.19°C in the to 1999. A high correlation coefficient of 0.79 on average ensemble mean (with a standard deviation of 0.43°C), and was obtained for the monthly mean discharge in the winter 2.61°C in MIROC, respectively. The difference of SAT season and indicated that the method is suitable for future around Japan (130E–140E, 30N–45N) between the 1990s climate change studies. and the 2070s is 2.53°C in the ensemble mean (with a The global warming numerical experiment suggested standard deviation of 0.70°C), and 3.36°C in MIROC, that the monthly mean hydrograph for the 2070s will differ respectively. The future global change of SAT in the MIROC from that observed in the 1990s. The monthly mean is larger than the ensemble mean (Figure S2 in Supplement). discharge was projected to increase by approximately 43% Therefore, this projection reflects a climate condition under in January and 55% in February, compared to the respective

—28— HYDROLOGICAL RESPONSE TO FUTURE CLIMATE CHANGE IN THE AGANO RIVER BASIN, JAPAN values for the 1990s. On the other hand, discharge will 471. decrease by approximately 38% in April and 32% in May, Kite GW, Haberlandt U. 1999. Atmospheric model data for respectively. Few changes in discharge were found for the macroscale hydrology. Journal of Hydrology 217: 303–313. period from July to December and in March. The 2070s doi:10.1016/S0022-1694(98)00230-3. flood peak was projected to occur one month earlier, in Kobayashi T, Kojiri T, Nozawa T. 2008. Estimation of precipitation variation in river basin scale with outputs from an atmosphere- March, compared to the April peak in the 1990s. It seems ocean general circulation model called by MIROC. Journal that the impact of rising temperatures due to the change in of Japan Society of Hydrology and Water Resources 21: 423– precipitation is limited, considering the lack of major 438. doi:10.3178/jjshwr.21.423 (in Japanese). changes in river discharge characteristics. Ma X, Fukushima Y, Hiyama T, Hashimoto T, Ohata T. 2000. A The PGW results also showed a marked change in macro-scale hydrological analysis of the Lena River basin. snowfall. The PGW snowfall amounts were 20% lower than Hydrological Processes 14(3): 639–651. doi: 10.1002/(SICI) those in the CTL-1990s, with a basin-wide average reduction 1099-1085(20000228)14:3<639::AID-HYP953>3.0.CO;2-0. of approximately 49.5% and a range from 23 to 95%. Ma X, Fukushima Y, Hashimoto T, Hiyama T, Nakashima T. 1999. Application of a simple SVAT model in a mountain catchment under temperate humid climate. Journal of Japan Society of ACKNOWLEDGEMENTS Hydrology and Water Resources 12: 285–294. Ministry of Land, Infrastructure, Transport and Tourism Hokuriku Regional Development Bureau. 2009. Agano River a la carte. We acknowledge the modeling groups, the Program for http://www.hrr.mlit.go.jp/agano/. [December 10, 2009] (in Climate Model Diagnosis and Intercomparison, and the Japanese). World Climate Research Programme (WCRP)’s Working Nash J, Sutcliffe J. 1970. River flow forecasting through conceptual Group on Coupled Modelling (WGCM) for their roles in models Part I—A discussion of principles. Journal of making the WCRP CMIP3 multi-model dataset available. Hydrology 10(3): 282–290. doi:10.1016/0022-1694 (70) We thank Dr. Hiroaki Kawase of the National Institute for 90255-6. Environmental Studies for preparing CMIP3 GCMs data set. Nozawa T, Nagashima T, Ogura T, Yokohata T, Okada N, Shiogama We also acknowledge helpful comments from the editor and H. 2007. Climate change simulations with a coupled ocean- atmosphere GCM called the model for interdisciplinary two anonymous reviewers. research on climate: MIROC. CGER’s Supercomputer Monograph Report 12: Center for Global Environmental REFERENCES Research, National Institute for Environmental Studies, Tsukuba, Japan; 79 pp. Skamarock WC, Klemp JB, Dudhia J, Gill DO, Barker DM, Wang Fujihara Y, Ode M, Kojiri T, Tomosugi K, Irie H. 2006. Effects W, Powers JG. 2005. A description of the advanced research of global warming on the water resources of the Tone River WRF version 2, NCAR Technical Note NCAR/ TN-468+STR. basin. Annual Journal of Hydraulic Engineering 50: 367–372 National Center for Atmospheric Research: Colorado, USA; (in Japanese). 88 pp. Hara M, Yoshikane T, Kawase H, Kimura F. 2008. Estimation of Tachikawa Y, Takino S, Ichikawa Y, Shiba M. 2009. Study on the the impact of global warming on snow depth in Japan by the impact of climate change on river flow regme in the Mogami Psedo-Global-Warming method. Hydrological Research and basins. Annual Journal of Hydraulic Letters 2: 61–64. doi: 10.3178/HRL.2.61. Engineering 53: 475–480 (in Japanese). IPCC. 2007. Climate Change 2007. The Physical Science Basis. U.S. Department of the Interior. GTOPO30. http://eros.usgs. Contribution of Working Group I to the Fourth Assessment gov/#/Find_Data/Products_and_Data_Available/gtopo30_info. Report of the Intergovernmental Panel on Climate Change, [December 10, 2009] Solomon S, Qin D, Manning M, Chen Z, Marquis M, Averyt Wada K, Kawasaki M, Tomizawa Y, Shoji Kusunoki S, Kazuo KB, Tignor M, Miller HL (eds.); Cambridge University Press: Kurihaea K. 2008. Study on the flood risk assessment of Cambridge, United Kingdom and New York, USA; 996 pp. global warming by regional climate model and high resolution Japan Meteorological Agency. 2002. Twenty century climate of general circulation model. Journal of Japan Society of Japan. The Ministry of Finance Printing Bureau: Tokyo; 116 Hydrology and Water Resources 21: 12–22 (in Japanese). pp. WMO (World Meteorological Organization). 1994. Guide to Kalnay E, Kanamitsu M, Kistler R, Collins W, Deaven D, Gandin Hydrological Practices (Fifth edition). WMO-164, WMO: L, Iredell M, Saha S, White G, Woollen J, Zhu Y, Leetmaa Geneva; 738 pp. A, Reynolds R, Chelliah M, Ebisuzaki W, Higgins W, Wood AW, Leung LR, Sridhar V, Lettenmaier DP. 2004. Hydrologic Janowiak J, Mo K, Ropelewski C, Wang J, Jenne R, Joseph implications of dynamical and statistical approaches to D. 1996. The NCEP/NCAR 40-Year Reanalysis Project. downscaling climate model outputs. Climate Change 62: 189– Bulletin of the American Meteorological Society 77(3): 437– 216. doi:10.1023/B:CLIM.0000013685.99609.9e.

—29—