Water cycle is essential substance move for human survival. Mathematical modeling of hydrology is the key tool to meet the more and more critical societal needs for improved water management and hazards prediction. The development of hydrological modeling has been in the direction from the first generation empirical and lumped models, to the second generation distributed hydrological models, and then to the third generation distributed biosphere-hydrological model. The new generation models incorporate the advanced schemes to understand the response of hydrological cycle to the change of biosphere, human society and climate system. The advances of methods for translating satellite data into global surface parameter sets have driven the development of the third generation models. With the new generation model, many of the merging advances in monitoring, computation, and telecommunications are brought to bear on disaster prevention, food supply, health, security and development issues facing Earth's growing population.

Multidisciplinary developments have prompted hydrological simulation. These advances include the new insights into mass/heat flux in the Soil-Vegetation-Atmosphere Transfer (SVAT) processes, progress in getting reliable land surface information from satellite remote sensing (RS), and developments of Geographic Information System (GIS) technique to extract topographic variables from digital elevation models (DEMs). This study has aimed at developing a new generation distributed biosphere-hydrological model accounting for the merging multidisciplinary advances. The main objectives include: extraction of hydrology related information from nontraditional datasets, development of a time-continuous distributed process-based land surface hydrological model accounting for the representations of the terrains, soil, vegetation, and hydrological response, model evaluation of the new generation model, and model application to evaluate the effects of human activity and natural climate variability on hydrology cycle.

The credibility of extraction of hydrology related information from nontraditional datasets is examined with a distributed biosphere-hydrological (DBH) model system. The nontraditional datasets to address water resource problems are largely from satellite remote sensing. The DBH model system is applied to the Yellow River basin, a continental scale river basin in semi-arid area, to compare the satellite remote sensing data with in situ observations. The relationship between nontraditional dataset and traditional in situ observation on cloud cover, which is characterized by large spatial and temporal variations, shows strong correlations, implying the credibility of use of nontraditional datasets in hydrological simulation. The DBH model system is then used to analyze hydro-climatic change and stream flow change, exploring the possible connections between hydrology, vegetation, climate and human activity in the Yellow River basin. The analysis, using station, satellite metadata and interpolated coverage, indicates that the precipitation decreases in most part of the Yellow River basin, climatic factors such as temperature and evaporative demand of the atmosphere have large trend in special part of the basin, and that human activities have changed the vegetation condition in the irrigation districts. The Loess Plateau, the Tibetan Plateau, and the irrigation districts are suggested as precipitation, temperature, and human activities "hot spots" of the Yellow River drying up, respectively.

A realistic distributed biosphere-hydrological (DBH) model has been developed for representing the role of both topography and land cover characteristics in hydrological cycle. The model is designed for use in a continental scale river basin and coupling with atmospheric models. The vegetated surface is calculated by a realistic land surface model SiB2. The vegetation phenology is described by satellite data, and the transfer of energy, mass and momentum between the atmosphere and land surface. The hydrological part estimates the surface runoff and calculates the interlayer exchanges within the soil profile and interaction between soil water and groundwater. The geomorphologic properties are abstracted from Digital Elevation Model using a distributed hydrological sub-model. Realistic watershed map and river way map are used to delineate sub-river basins and river network. The sub-river basins are coded following a natural numbering scheme which is self-replicating and is possible to provide identification numbers to the level of the smallest sub-basins. The river network routing order of the sub-basins is implicated in the numbering scheme. The runoff is then accumulated and routed to outlet using kinematic wave approach. The parameters and forcing data are obtained from various ways, including remote sensing, ground observation, and statistical surveys. The hydrology-related information was digitized into the model system in order to diminish the uncertainty in the hydrological simulation.

The model evaluation processes, such as model verification, validation and credibility, are preformed in the Yellow River basin, China. The effects of natural and anthropogenic heterogeneity on hydrological simulation were evaluated using the DBH model system. The model system embeds a biosphere model into distributed hydrological scheme, representing both topography and vegetation conditions in mesoscale hydrological simulation. An irrigation scheme has been included in the model system. The effects on hydrological processes of two kinds of variability, precipitation variability and the variability on irrigation redistributing runoff, was investigated in this study, representing the natural and anthropogenic heterogeneity, respectively. Runoff is underestimated if the rainfall is spatially uniformly put over large grid cell. And runoff simulation could be improved by taking into account the precipitation heterogeneity. However, the negative runoff contribution cannot be simulated by only considering the natural heterogeneity. This constructive model shortcoming can be eliminated by taking into account anthropogenic heterogeneity, irrigation water withdrawals. Irrigation leads to increased evapotranspiration and decreased runoff. Surface soil moisture in the irrigated area increases because of irrigation. Simulations performed for the Yellow River basin indicates that stream flow decrease of 41% by irrigation.　The latent heat flux increase in peak irrigation season (June, July, August: JJA) is 3.3 Wm(-2) with a decrease in ground surface temperature of 0.1 K of the river basin. The maximum simulated increase in latent heat flux is 43 Wm(-2) and ground temperature decrease is 1.6 K in peak irrigation season (JJA).

A comprehensive application of the DBH model system is performed in the Yellow River basin with the use of data analysis results to evaluate the effects of human activity and natural climate variability on hydrology cycle. Scenario simulations are performed from 1960 to 2000 to quantify the effects of human activity on hydrology, and to distinguish it from the effects from natural climate variability. The linear tendency of the forcing data is removed to provide input for non-change scenarios. The model results from six scenarios, i.e. most realistic control scenario, non-climate change scenario, non-vegetation change scenario, non-irrigated area change scenario, stable scenario without linear tendency and stable scenario without climate pattern change, are compared. The results show that climate change is dominated in the upper and middle reaches, but human activities are dominated in the lower reaches of the Yellow River basin. The runoff and evapotranspiration decrease over the Loess Plateau is dominated by the contribution of climate change. The intensively affected area by irrigation and vegetation change is the irrigation districts especial in the Weihe irrigation district and lower reaches irrigation districts. The river discharge at the river mouth nearly half is affected by climate change and half by human activities. The linear climate change contributes to the water consumption, but the climate pattern change is more important than the linear climate change. The river channel flow is more significant affected by the direct irrigation water withdrawals than by the climate change, which dominantly contributes to the annual water resources change. The reservoirs are believed to make more stream flow consumption for irrigation, at the same time, our results demonstrate the reservoirs help to keep environment flow and counter zero-flow in the river channel.

　水循環は人類を始めとする生物を支える根源である。健全な水循環の保全、水供給の安定は、健康や農業生産、工業生産などを通じて社会の持続的な発展、安定と平和に寄与する。災害を回避しつつ水循環の恩恵を適切に受けるためには、水循環の現状と将来予測とを知ることが不可欠であり、水循環の数値モデルは将来予測をして統合的な水資源管理をするために有用である。

　本論文の2章で詳しく述べられている通り、水循環の数値モデルは、入力としての降雨量もしくは融雪量と出力としての流量とを対応付けるモデルであったが、近年では降雨等の入力や土壌水分などの予報変数が水平分布を持つような分布型水文モデルも広く利用されるようになってきている。さらに、地表面の水やエネルギーの収支や循環に加えて、植生がそうしたエネルギー水循環に及ぼす影響を陽に表現した数値モデルが開発されるようになってきている。

　また、自然の水循環のみならず、土地利用の変化や地形改変、人工的貯水池による貯留と放流、河川からの灌漑取水や地下水くみ上げなど人間活動が水循環に及ぼしている影響も積極的に取り込んでモデルで表現し、人間活動による水循環の改変がグローバルスケールでさえ大きな影響を持っている現実の水循環をきちんとシミュレートする研究にも先端的に取り組まれている。

　本論文では、そうした植生の影響を考慮した水・エネルギー収支を算定可能な1次元の陸面モデルを2次元化し、分布型生物圏(DBH)水文モデルシステムのコアとして利用し、これに、モデルパラメータや外力、人間活動を考慮するサブシステムを結合して全体として植生の効果や人間活動の影響を考慮しつつ現実の水循環をシミュレート可能なDBHモデルを構築し、検証し、中国・黄河における長期の流量変動の要因の分析に応用している。

　第1章では研究の動機として世界の水問題、特に流量減少、極端な場合には断流として知られるような極端な流量低下問題に触れ、そうした問題に立ち向かう学問としての水文科学に求められている研究のあり方、そして本研究の目標と構成について紹介されている。

　第2章ではいわゆる水文モデルと陸面モデルとの歴史的な開発の経緯と世界的な発展段階の状況について広範なレビューととりまとめを行い、分布型水文モデルと植生を考慮した陸面モデルとを結合したDBHモデルシステムの開発という本研究の位置づけがなされている。

　第3章ではDBHモデルへの外力データを整備するにあたり、従来からの地点データの空間内挿法以外に、衛星リモートセンシングデータに基づいた葉面積指数や光合成有効放射量データの利用、さらには、正規化植生指数(NDVI)の時系列から雲量を逆推定する手法を開発してその推定精度を検証した上で、外力として利用している。さらには、長期観測データに基づき、基本気象要素についての長期トレンドを解析し、最後の応用の章に繋がる成果を得ている。

　第4章では、DBHの構築に関して記述されている。鉛直1次元の水エネルギー収支はSiB2と呼ばれる陸面モデルによって算定され、その入力には10kmといった比較的大きな水平グリッドの内部での降水強度の不均一性も考慮されている。表面流出、地下水流出は勾配を考慮しつつキネマティックウエイブ近似で河道に流下させ、地下水と河道内の水との交換も考慮されている。また、灌漑スキームや物理パラメータの与え方に関しても説明されている。

　第5章では、黄河に適用した結果が示され、構築されたDBHシステムが過去の流出量の変動を的確に反映できること、また、降水強度の空間不均一性の考慮が及ぼす影響、灌漑取水が及ぼす影響に関する感度分析の結果も示されている。

　第6章では1960年から2000年の中国黄河の流量シミュレーションにおいて、長期変動がどういう要因にもたらされたかを調べる数値実験が行われた。その結果、黄河上流、黄土高原における蒸発散量の減少に対しては気候変動の影響が大きいが、下流域では灌漑取水量の増加など人間活動が長期変動要因として支配的であることが示されている。その結果、河口付近では人間活動影響の方が気候変動の影響よりも強いことがわかった。こうした結果により、貯留施設が生態系維持のための環境用水の保全や断流の阻止に有効であることが示されている。

　以上、本研究は、植生の影響を陽に取り込んだ分布型水文モデルシステムを構築し、それが流域の水循環の長期的な変動を適切に表現できるのみならず、黄河断流の要因分析と解決策の検討といった重要な社会的課題の解決へ向けた取り組みにも応用可能であることを示した画期的なものであり、有用性に富む独創的な研究成果と評価できる。よって本論文は博士(工学)の学位請求論文として合格と認められる。