京都大学 / 0048 / 新制・課程博士 / 博士(工学) / 甲第17876号 / 工博第3785号 / 新制||工||1579(附属図書館) / 30696 / 京都大学大学院工学研究科都市社会工学専攻 / (主査)教授 角 哲也, 教授 堀 智晴, 准教授 田中 賢治, 准教授 Sameh Ahmed Kantoush / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DGAM

During the 21st century, climatic change and non-climatic stressors are likely to impact forests leading to large-scale forest and biodiversity loss, and diminished ecological benefits. Assessing the vulnerability of forests and addressing the sources of vulnerability is an important risk management strategy. The overall goal of this research work is to develop methodological approaches at different scales and apply them to assess the vulnerability of forests in India for developing strategies for forest adaptation. Indicator-based methodological approaches have been developed for vulnerability assessment at local, landscape and national scales under current climate scenario, and at national scale under future climate scenario. Under current climate scenario, the concept of inherent vulnerability of forests has emerged by treating vulnerability as a characteristic internal property of a forest ecosystem independent of exposure. This approach to assess vulnerability is consistent with the framework presented in the latest report of Intergovernmental Panel on Climate Change (IPCC AR5 2014). Assessment of vulnerability under future climate scenario is presented only at national scale due to challenges associated with model-based climate projections and impact assessment at finer scales. The framework to assess inherent vulnerability of forests at local scale involves selection of vulnerability indicators and pair wise comparison method (PCM) to assign the indicator weights. The methodology is applied in the field to a 300-ha moist deciduous case study forest (Aduvalli Protected Forest, Chikmagalur district) in the Western Ghats area, where a vulnerability index value of 0.248 is estimated. Results of the study indicate that two indicators - ‘preponderance of invasive species’ and ‘forest dependence of community’ - are the major drivers of inherent vulnerability at present. The methodology developed to assess the inherent vulnerability at landscape scale involves use of vulnerability indicators, the pair wise comparison method, and geographic information system (GIS) tools. Using the methodology, assessment of inherent vulnerability of Western Ghats Karnataka (WGK) landscape forests is carried out. Four vulnerability indicators namely, biological richness, disturbance index, canopy cover and slope having weights 0.552, 0.266, 0.123 and 0.059, respectively are used. The study shows that forests at one-third of the grid points in the landscape have high and very high inherent vulnerability, and natural forests are inherently less vulnerable than plantation forests. The methodology used for assessment of forest inherent vulnerability at the national scale was same as used at landscape scale. 40% of forest grid points in India are assessed with high and very high inherent vulnerability. Except in pockets, the forests in the three biodiversity hotspots in India i.e., the Western Ghats in peninsular India, northeastern India, and the northern Himalayan region are assessed to have low to medium inherent vulnerability. Vulnerability of forests under future climate scenario at national scale is estimated by combining the results of assessment of climate change impact and inherent vulnerability. In the present study, ensemble climatology from five CMIP5 (Coupled Model Intercomparison Project phase 5) climate models for RCP (Representative Concentration Pathways) 4.5 and 8.5 in short (2030s) and long term (2080s) is used as input to IBIS (Integrated Biosphere Simulator) dynamic vegetation model. Forest grid points projected to experience vegetation-shift to a new plant functional type (PFT) under future climate are categorized under ‘extremely high’ vulnerability category. Such forest grid points in India are 22 and 23% in the short term under RCP4.5 and 8.5 respectively, and these percentages increase to 31 and 37% in the long term. IBIS simulated vegetation projections are also compared with LPJ (Lund-Potsdam-Jena) simulated projections. Both the vegetation models agree that forests at about one-third of the grid points could be impacted by future climate but the spatial distribution of impacted grid points differs between the models. Vulnerability assessment is a powerful tool for building long-term resilience in the forest sector in the context of projected climate change. From this study, three forest scenarios emerge in India for developing adaptation strategies namely: (a) less disturbed primary forests; (b) degraded and fragmented primary forests; and (c) secondary (plantation) forests. Minimizing anthropogenic disturbance and conserving biodiversity are critical to reduce forest vulnerability of less disturbed primary forests. For disturbed forests and plantations, adaptive management aimed at forest restoration is necessary to build resilience. Mainstreaming forest adaptation in India through Forest Working Plans and realignment of the forestry programs is necessary to manage the risk to forests under climate change.

<p>This dissertation examines the impact of the expansion of renewable technology at both national and local level, through distinct essays. At the national level, the first paper analyzes the effects of economic and distributional impacts of climate mitigation policy, in the context of a developing country, to understand the interactions between the energy system and the macroeconomic environment. In the case of the local level, the second paper uses synthetic control method, to estimate the effect at the county level of utility scale wind in the development indicators for two counties in the U.S. </p> <p>The first paper assesses the economic and distributional impacts of Nicaragua’s commitments to limit future greenhouse gas emissions in the context of the Paris Agreement, known as the Nationally Determined Contributions (NDCs). The analysis relies on two distinct models. The first is a top-down approach based on a single-country computable general equilibrium (CGE) model, known as the Mitigation, Adaptation and New Technologies Applied General Equilibrium (MANAGE) Model. The second is a bottom-up approach based on the Open-Source energy Modeling System (OSeMOSYS), which is technology rich energy model. The combined model is calibrated to an updated social accounting matrix for Nicaragua, which disaggregates households into 20 representative types: 10 rural and 10 urban households. For the household disaggregation we have used information from the 2014 Living Standards Measurement Study (LSMS) for Nicaragua. Our analysis focuses on the distributional impacts of meeting the NDCs as well as additional scenarios—in a dynamic framework as the MANAGE model is a (recursive) dynamic model. The results show that a carbon tax has greatest potential for reduction in emissions, with modest impact in macro variables. An expansion of the renewable sources in the electricity matrix also leads to significant reduction in emissions. Only a carbon tax achieves a reduction in emissions consistent with keeping global warming below 2°C. Nicaragua’s NDC alone would not achieve the target and mitigation instruments are needed. An expansion of generation from renewable sources, does not lead to a scenario consistent with a 2°C pathway. </p> <p>The second paper measures the impact of wind generation on county level outcomes through the use of the Synthetic Control Method (SCM). SCM avoids the pitfalls of other methods such as input-output models and project level case studies that do not provide county level estimates. We find that the local per capita income effect of utility wind scale is 6 percent (translate into an increase of $1,511 in per capita income for 2019) for Benton County and 8 percent for White county in Indiana (an increase of $2,100 in per capita income for 2019). The per capita income effect measures the average impact, which includes the gains in rents from capital, land, and labor from wind power in these counties. Moreover, we find that most of the rents from wind power accrue to the owners of capital and labor. Even assuming the lowest projections of electricity prices and the highest reasonable cost we still find a 10 percent minimum rate of return to capital for both Benton and White counties’ wind power generators. Furthermore, we find that there are excess rents that could be taxed and redistributed at the county, state, or federal level without disincentivizing investment in wind power.</p>

Climate change could aggravate periodic and chronic shortfalls of water, particularly in arid and semi-arid areas of the world (IPCC, 2001). Climate change is likely to accelerate the global hydrological cycle, with increase in temperature, changes in precipitation patterns, and evapotranspiration affecting the water quantity and quality, water availability and demands. The various components of a surface water resources system affected by climate change may include the water availability, irrigation demands, water quality, hydropower generation, ground water recharge, soil moisture etc. It is prudent to examine the anticipated impacts of climate change on these different components individually or combinedly with a view to developing responses to minimize the climate change induced risk in water resources systems. Assessment of climate change impacts on water resources essentially involves downscaling the projections of climatic variables (e.g., temperature, humidity, mean sea level pressure etc.) to hydrologic variables (e.g., precipitation and streamflow), at regional scale. Statistical downscaling methods are generally used in the hydrological impact assessment studies for downscaling climate projections provided by the General Circulation Models (GCMs). GCMs are climate models designed to simulate time series of climate variables globally, accounting for the greenhouse gases in the atmosphere. The statistical techniques used to bridge the spatial and temporal resolution gaps between what GCMs are currently able to provide and what impact assessment studies require are called as statistical downscaling methods. Generally, these methods involve deriving empirical relationships that transform large-scale simulations of climate variables (referred as the predictors) provided by a GCM to regional scale hydrologic variables (referred as the predictands). This general methodology is characterized by various uncertainties such as GCM and scenario uncertainty, uncertainty due to initial conditions of the GCMs, uncertainty due to downscaling methods, uncertainty due to hydrological model used for impact assessment and uncertainty resulting from multiple stake holders in a water resources system. The research reported in this thesis contributes towards (i) development of methodologies for climate change impact assessment of various components of a water resources system, such as water quality, water availability, irrigation and reservoir operation, and (ii) quantification of GCM and scenario uncertainties in hydrologic impacts of climate change. Further, an integrated reservoir operation model is developed to derive optimal operating policies under the projected scenarios of water availability, irrigation water demands, and water quality due to climate change accounting for various sources of uncertainties. Hydropower generation is also one of the objectives in the reservoir operation. The possible climate change impact on river water quality is initially analyzed with respect to hypothetical scenarios of temperature and streamflow, which are affected by changes in precipitation and air temperature respectively. These possible hypothetical scenarios are constructed for the streamflow and river water temperature based on recent changes in the observed data. The water quality response is simulated, both for the present conditions and for conditions resulting from the hypothetical scenarios, using the water quality simulation model, QUAL2K. A Fuzzy Waste Load Allocation Model (FWLAM) is used as a river water quality management model to derive optimal treatment levels for the dischargers in response to the hypothetical scenarios of streamflow and water temperature. The scenarios considered for possible changes in air temperature (+1 oC and +2 oC) and streamflow (-0%, -10%, -20%) resulted in a substantial decrease in the Dissolved Oxygen (DO) levels, increase in Biochemical Oxygen Demand (BOD) and river water temperature for the case study of the Tunga-Bhadra River, India. The river water quality indicators are analyzed for the hypothetical scenarios when the BOD of the effluent discharges is at safe permissible level set by Pollution Control Boards (PCBs). A significant impairment in the water quality is observed for the case study, under the hypothetical scenarios considered. A multi-variable statistical downscaling model based on Canonical Correlation Analysis (CCA) is then developed to downscale future projections of hydro¬meteorological variables to be used in the impact assessment study of river water quality. The CCA downscaling model is used to relate the surface-based observations and atmospheric variables to obtain the simultaneous projection of hydrometeorological variables. Statistical relationships in terms of canonical regression equations are obtained for each of the hydro-meteorological predictands using the reanalysis data and surface observations. The reanalysis data provided by National Center for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) are used for the purpose. The regression equations are applied to the simulated GCM output to model future projections of hydro-meteorological predictands. An advantage of the CCA methodology in the context of downscaling is that the relationships between climate variables and the surface hydrologic variables are simultaneously expressed, by retaining the explained variance between the two sets. The CCA method is used to model the monthly hydro-meteorological variables in the Tunga-Bhadra river basin for water quality impact assessment study. A modeling framework of risk assessment is developed to integrate the hydro¬meteorological projections downscaled from CCA model with a river water quality management model to quantify the future expected risk of low water quality under climate change. A Multiple Logistic Regression (MLR) is used to quantify the risk of Low Water Quality (LWQ) corresponding to a threshold DO level, by considering the streamflow and water temperature as explanatory variables. An Imprecise Fuzzy Waste Load Allocation Model (IFWLAM) is adopted to evaluate the future fractional removal policies for each of the dischargers by including the predicted future risk levels. The hydro-meteorological projections of streamflow, air temperature, relative humidity and wind speed are modeled using MIROC 3.2 GCM simulations with A1B scenario. The river water temperature is modeled by using an analytical temperature model that includes the downscaled hydro-meteorological variables. The river water temperature is projected to increase under climate change, for the scenario considered. The IFWLAM uses the downscaled projections of streamflow, simulated river water temperature and the predicted lower and upper future risk levels to determine the fraction removal policies for each of the dischargers. The results indicate that the optimal fractional removal levels required for the future scenarios will be higher compared to the present levels, even if the effluent loadings remain unchanged. Climate change is likely to impact the agricultural sector directly with changes in rainfall and evapotranspiration. The regional climate change impacts on irrigation water demands are studied by quantifying the crop water demands for the possible changes of rainfall and evapotranspiration. The future projections of various meteorological variables affecting the irrigation demand are downscaled using CCA downscaling model with MIROC 3.2 GCM output for the A1B scenario. The future evapotranspiration is obtained using the Penman-Monteith evapotranspiration model accounting for the projected changes in temperature, relative humidity, solar radiation and wind speed. The monthly irrigation water demands of paddy, sugarcane, permanent garden and semidry crops quantified at nine downscaling locations covering the entire command area of the Bhadra river basin, used as a case study, are projected to increase for the future scenarios of 2020-2044, 2045-2069 and 2070-2095 under the climate change scenario considered. The GCM and scenario uncertainty is modeled combinedly by deriving a multimodel weighted mean by assigning weights to each GCM and scenario. An entropy objective weighting scheme is proposed which exploits the information contained in various GCMs and scenarios in simulating the current and future climatology. Three GCMs, viz., CGCM2 (Meteorological Research Institute, Japan), MIROC3.2 medium resolution (Center for Climate System Research, Japan), and GISS model E20/Russell (NASA Goddard Institute for Space Studies, USA) with three scenarios A1B, A2 and B1 are used for obtaining the hydro-meteorological projections for the Bhadra river basin. Entropy weights are assigned to each GCM and scenario based on the performance of the GCM and scenario in reproducing the present climatology and deviation of each from the projected ensemble average. The proposed entropy weighting method is applied to projections of the hydro-meteorological variables obtained based on CCA downscaling method from outputs of the three GCMs and the three scenarios. The multimodel weighted mean projections are obtained for the future time slice of 2020-2060. Such weighted mean hydro-meteorological projections may be further used into the impact assessment model to address the climate model uncertainty in the water resources systems. An integrated reservoir operation model is developed considering the objectives of irrigation, hydropower and downstream water quality under uncertainty due to climate change, uncertainty introduced by fuzziness in the goals of stakeholders and uncertainty due to the random nature of streamflow. The climate model uncertainty originating from the mismatch between projections from various GCMs under different scenarios is considered as first level of uncertainty, which is modeled by using the weighted mean hydro-meteorological projections. The second level of uncertainty considered is due to the imprecision and conflicting goals of the reservoir users, which is modeled by using fuzzy set theory. A Water Quantity Control Model (WQCM) is developed with fuzzy goals of the reservoir users to obtain water allocations among the different users of the reservoir corresponding to the projected demands. The water allocation model is updated to account for the projected demands in terms of revised fuzzy membership functions under climate change to develop optimal policies of the reservoir for future scenarios. The third level of uncertainty arises from the inherent variability of the reservoir inflow leading to uncertainty due to randomness, which is modeled by considering the reservoir inflow as a stochastic variable. The optimal monthly operating polices are derived using Stochastic Dynamic Programming (SDP), separately for the current and for the future periods of 2020-2040 and 2040-2060 The performance measures for Bhadra reservoir in terms of reliability and deficit ratios for each reservoir user (irrigation, hydropower and downstream water quality) are estimated with optimal SDP policy derived for current and future periods. The reliability with respect to irrigation, downstream water quality and hydropower show a decrease for 2020-2040 and 2040-2060, while deficit ratio increases for these periods. The results reveal that climate change is likely to affect the reservoir performance significantly and changes in the reservoir operation for the future scenarios is unable to restore the past performance levels. Hence, development of adaptive responses to mitigate the effects of climate change is vital to improve the overall reservoir performance.

<p>Forest management involves numerous stochastic elements. To sustainably manage forest resources, it is crucial to acknowledge these sources as uncertainty or risk, and incorporate them in adaptive decision-making. Here, I developed several stochastic programming models in the form of passive or active adaptive management for natural mixed-species hardwood forests in Indiana. I demonstrated how to use these tools to deal with time-invariant and time-variant natural disturbances in optimal planning of harvests.</p> <p> Markov decision process (MDP) models were first constructed based upon stochastic simulations of an empirical forest growth model for the forest type of interest. Then, they were optimized to seek the optimal or near-optimal harvesting decisions while considering risk and uncertainty in natural disturbances. In particular, a classic expected-criterion infinite-horizon MDP model was first used as a passive adaptive management tool to determine the optimal action for a specific forest state when the probabilities of forest transition remained constant over time. Next, a two-stage non-stationary MDP model combined with a rolling-horizon heuristic was developed, which allowed information update and then adjustments of decisions accordingly. It was used to determine active adaptive harvesting decisions for a three-decade planning horizon during which natural disturbance probabilities may be altered by climate change.</p> <p> The empirical results can be used to make some useful quantitative management recommendations, and shed light on the impacts of decision-making on the forests and timber yield when some stochastic elements in forest management changed. In general, the increase in the likelihood of damages by natural disturbance to forests would cause more aggressive decisions if timber production was the management objective. When windthrow did not pose a threat to mixed hardwood forests, the average optimal yield of sawtimber was estimated to be 1,376 ft³/ac/acre, while the residual basal area was 88 ft²/ac. Assuming a 10 percent per decade probability of windthrow that would reduce the stand basal area considerably, the optimal sawtimber yield per decade would decline by 17%, but the residual basal area would be lowered only by 5%. Assuming that the frequency of windthrow increased in the magnitude of 5% every decade under climate change, the average sawtimber yield would be reduced by 31%, with an average residual basal area slightly around 76 ft²/ac. For validation purpose, I compared the total sawtimber yield in three decades obtained from the heuristic approach to that of a three-decade MDP model making <i>ex post</i> decisions. The heuristic approach was proved to provide a satisfactory result which was only about 18% lower than the actual optimum.</p> These findings highlight the need for landowners, both private and public, to monitor forests frequently and use flexible planning approaches in order to anticipate for climate change impacts. They also suggest that climate change may considerably lower sawtimber yield, causing a concerning decline in the timber supply in Indiana. Future improvements of the approaches used here are recommended, including addressing the changing stumpage market condition and developing a more flexible rolling-horizon heuristic approach.