It has been hypothesized that local or global climatic conditions can affect societies in a variety of ways. However, to date, it has been difficult to measure the social impact of climate, so the scale and scope of its influence on populations remains mostly theoretical. This dissertation integrates data and quantitative methods from climate science, economics and political science to develop new techniques for empirically measuring the the social impacts of climate. These techniques are used to measure large-scale dynamical relationships between climatological conditions and the response of the societies that are exposed to them. In general, the response of societies to climatological forcing is found to be larger than previously thought. The concluding chapter discusses how these findings may inform policies that govern the global environment and economic development.

This dissertation analyzes three di_erent dimensions of climate risk: (i) impacts and responses to climate change in physical and biological systems; (ii) socio-economic consequences of climatic variability in human systems; and (iii) the design of formal insurance instruments to reduce socio-economic vulnerability to climatic risk, as adaptation strategies. Each part represents an independent study.

Warming of both the high latitudes and tropical sea surface temperatures are present in modern observations and projected under future climate change scenarios. These conditions were also present in the Warm Pliocene (3.3 - 3.0 million years ago), a paleoclimatic interval that bares resemblance to future global warming. This dissertation investigates the impact of both tropical and high latitude warming on regional atmospheric circulation using GISS global climate model simulations of the Pliocene and sensitivity tests. Chapter 1 discusses the initial approach used to investigate how a warmer climate impacts regional atmospheric circulation. A general circulation model (GCM) was utilized to assess the contribution from both high latitude and tropical warming to regional Pliocene climatic patterns. It was found that both a warming of the high latitudes and Indo-Pacific tropical region are needed to reproduce the regional Pliocene climates indicated by terrestrial paleo-proxy data. These results suggest that the tropical atmospheric circulation of the Indo-Pacific region during the warm Pliocene may have been different from modern mean conditions. These findings are corroborated by Pliocene paleo-data, a luxury not afforded by future climate projections, and provide insight into possible regional atmospheric circulation processes in a future warmer climate. Chapter 2 (Shukla et al., 2011) investigates how exactly the Indo-Pacific circulation and global teleconnections differed from modern day conditions. GCM generated teleconnections from the Indo-Pacific region were examined from origin to their impact on the extra-tropics under warm Pliocene conditions. The exact forcing source was not assumed a-priori, and it was found that while warmer SSTs in the eastern tropical Pacific generated weak El Niño-like teleconnections to North America, their effects over the Indian Ocean region were attenuated, primarily by the warmer SSTs there. Teleconnections to the extra-tropics were largely blocked from the Indian Ocean region, and most of the energy generated by the SST patterns went into maintaining an anomalous atmospheric overturning circulation. This altered background circulation of the Indian Ocean region can impact the South Asian Summer Monsoon (SASM) system. In these simulations, the dynamic monsoon intensity experienced the greatest decrease with tropical warming alone. Lesser SASM weakening occurred when both tropical and high latitude warming were imposed. Given the potential Indo-Pacific SSTs changes under Pliocene and warm climate conditions, Chapters 3 and 4 focus on the implications these changes have for the South Asian Summer Monsoon circulation. Chapter 3 examines the GISS suite of GCMs' ability to reproduce the major features of the South Asian Summer Monsoon (SASM) system. The GISS Model E (atmosphere only), Middle Atmospheres Model 3 (atmosphere only) and the ocean-atmosphere coupled Model E were run using forcings from 1960-2008. Major indices and features of the SASM were evaluated and compared to NCEP/NCAR and ECMWF reanalysis data. It was found that the atmosphere-only Model E better simulated, both in magnitude and variability, the circulatory (wind, vorticity, etc.) components of the SASM, whereas the coupled ModelE better simulated the magnitude of rainfall over the Indian sub-continent. Chapter 3 highlighted the SASM features in the models that need improvement, specifically in the overproduction of rainfall and the underestimation of windspeeds. Given the relatively accurately modelE simulated SASM intensity variability, and acknowledging its underestimation of wind strength, continuing modelE studies of the SASM will focus on large-scale circulation processes, rather than the rainfall distribution and variability. Chapter 4 compares SASM changes under both Pliocene conditions and future climate projections, the latter dictated by the Representative Concentration Pathways (RCPs). A tropical SST forcing, in the form of warmer western tropical Indian Ocean and eastern tropical Pacific Ocean SSTs, was additionally tested in isolation from globally warmer conditions. It was found that the SASM weakens under globally warmer conditions, but the greatest weakening occurred under tropical forcing alone. This suggests the importance of the relative regional temperature gradients of the Indian Ocean region. Although both simulations served to weaken the SASM system, the regional climatic patterns differed between Pliocene and future simulations and warrant further investigation. Future studies must focus on obtaining more data from the Indian Ocean region for the Pliocene period in order to corroborate modeled climate processes in that region. In addition, more assessments must be done to understand difference between climate processes in future projections and past warm climate intervals as this will aid in model development and our understanding of the climate's response and sensitivity.

Tropical cyclones (TC) in the western North Pacific (WNP) pose a serious threat to the coastal regions of Eastern Asia when they make landfall. The limited amount of observational data and the high computational cost of running TC-permitting dynamical models indicate a need for statistical models that can simulate large ensembles of TCs in order to cover the full range of possible activity that results from a given climate change. I construct and apply a statistical track model from the 1945-2007 observed "best tracks" in the IBTrACS database for the WNP. The lifecycle components--genesis, track propagation, and death--of each simulated track is determined stochastically based on the statistics of historical occurrences. The length scale that dictates what historical data to consider as "local" for each lifecycle component is calculated objectively through optimization. Overall, I demonstrate how a statistical model can be used as a tool to translate climate-induced changes in TC activity into implications for risk. In contrast to other studies, I show that the El Niño/Southern Oscillation (ENSO) has an effect on track propagation separate from the genesis effect. The ENSO effect on genesis results in higher landfall rates during La Niña years due to the shift in genesis location to the northeast. The effect on tracks is more geographically and seasonally varied due to local changes in the mid-level winds. I use local regression techniques to capture the relationship between ENSO, cyclogenesis, and track propagation. Stationary climate simulations are run for extreme ENSO states in order to better understand changes in TC activity and their implication for regional landfall. Additionally, Several diagnostics are performed on model realizations of the historical period, confirming the model's ability to capture the geographical distribution and interannual variability of observed TCs. Lastly, as a step to connect synthetic TC track simulations to economic damage risk assessment, I use a Damage Index and total damage data for U.S. landfalling hurricanes and fit generalized Pareto distributions (GPD) to them. The Damage Index uniquely separates out the effects of the physical damage capacity of a TC and the local economic vulnerability of a coastal region. GPD fits are also performed using covariates in the scale parameter, where bathymetric slope and landfall intensity are found to be useful covariates for the Damage Index. Using the Damage Index with covariates model, two examples are shown of assessing damage risk for different climates. The first takes landfall data input from a statistical-deterministic TC model downscaled from GFDL and ECHAM model current and future climates. The second takes landfall data from a fully statistical track model with different values of relative sea surface temperature given as a statistical predictor.

Tropical cyclones (TC) in the western North Pacific (WNP) pose a serious threat to the coastal regions of Eastern Asia when they make landfall. The limited amount of observational data and the high computational cost of running TC-permitting dynamical models indicate a need for statistical models that can simulate large ensembles of TCs in order to cover the full range of possible activity that results from a given climate change. I construct and apply a statistical track model from the 1945-2007 observed "best tracks" in the IBTrACS database for the WNP. The lifecycle components--genesis, track propagation, and death--of each simulated track is determined stochastically based on the statistics of historical occurrences. The length scale that dictates what historical data to consider as "local" for each lifecycle component is calculated objectively through optimization. Overall, I demonstrate how a statistical model can be used as a tool to translate climate-induced changes in TC activity into implications for risk.In contrast to other studies, I show that the El Niño/Southern Oscillation (ENSO) has an effect on track propagation separate from the genesis effect. The ENSO effect on genesis results in higher landfall rates during La Niña years due to the shift in genesis location to the northeast. The effect on tracks is more geographically and seasonally varied due to local changes in the mid-level winds. I use local regression techniques to capture the relationship between ENSO, cyclogenesis, and track propagation. Stationary climate simulations are run for extreme ENSO states in order to better understand changes in TC activity and their implication for regional landfall. Additionally, Several diagnostics are performed on model realizations of the historical period, confirming the model's ability to capture the geographical distribution and interannual variability of observed TCs. Lastly, as a step to connect synthetic TC track simulations to economic damage risk assessment, I use a Damage Index and total damage data for U.S. landfalling hurricanes and fit generalized Pareto distributions (GPD) to them. The Damage Index uniquely separates out the effects of the physical damage capacity of a TC and the local economic vulnerability of a coastal region. GPD fits are also performed using covariates in the scale parameter, where bathymetric slope and landfall intensity are found to be useful covariates for the Damage Index. Using the Damage Index with covariates model, two examples are shown of assessing damage risk for different climates. The first takes landfall data input from a statistical-deterministic TC model downscaled from GFDL and ECHAM model current and future climates. The second takes landfall data from a fully statistical track model with different values of relative sea surface temperature given as a statistical predictor.

This thesis seeks an improved understanding of what has been driving changes in the large scale tropospheric circulation. First, we consider the effects of stratospheric water vapor levels, which exhibit significant changes on both interannual and decadal timescales. It is shown that idealized thermal forcings mimicking increases in stratospheric water vapor produce poleward expansion of the Hadley cells (HCs) and poleward shifts of the midlatitude jets. Quantitatively, the circulation responses are comparable to those produced by increased well-mixed greenhouse gases. This suggests that stratospheric water vapor may be a significant contribution to past and projected changes in the tropospheric circulation. The second part of this thesis focuses on the response to idealized thermal forcings in the troposphere. It is found that zonally uniform warming confined to a narrow region around the equator produces contraction of the HCs and equatorward shifts of the midlatitude jets. Forcings with wider meridional extent produce the opposite effect: HC expansion and poleward shifts of the jets. If the forcing is confined to the midlatitudes, the amount of HC expansion is more than three times that of a forcing of comparable amplitude that is spread over the tropics. This finding may be relevant to recently observed trends of amplified warming in the midlatitudes. Furthermore, a simple diffusive model is constructed to explain the sensitivity of the circulation response to the meridional structure of the thermal forcing. The final part of this thesis considers the possible influence of solar forcing on the tropospheric circulation. Of particular interest is the steady state response to a 0.1% increase in total solar irradiance (TSI), the approximate amplitude of the 11-year solar cycle. Using a comprehensive atmospheric general circulation model coupled to a mixed layer ocean, it is found that a 0.1% TSI increase produces a circulation response that has a high dependence on the background state. Specifically, a TSI perturbation applied to a present day climate produces an equatorward shift of the Southern Hemisphere (SH) midlatitude jet, while the same forcing applied to a warmer climate produces a poleward shift of the SH jet. Opposite-signed responses are also evident in regions of the sea surface temperature, sea level pressure, and precipitation fields. These divergent responses may help to explain why earlier studies reach highly disparate conclusions about the influence of solar variations on climate.

In this dissertation I focus my analyses of adaptation strategies to climate change on two areas of primary concern: (i) high-altitude ecosystems of the Tropical Andes, with particular interest in the so-called páramo ecosystems; and (ii) mosquito-borne diseases, focusing on Plasmodium falciparum malaria infections. My research on páramo ecosystems follows a six-tiered approach to understand the linkages between the ongoing changes in climatic conditions and the disruptions affecting the integrity of high-altitude environments. Activities conducted herein include the analyses of changes in atmospheric stability and lifting condensation levels; the diagnosis of changes in hydrological regimes; the assessment of the extent of life zones; the analyses of increases in the occurrence and rapid spread of high-altitude fires; the assessment of the integrity of páramo ecosystems; and the analyses of increases in climatic stress. Activities are conducted for three key, strategic, protected high-altitude Andean environments of the Tropical Andes and for the full length of the Andes Cordillera. My research findings provide elements for an improved understanding of the potential responses of Andean ecosystems to the large-scale rapidly changing climate and to a strongly-influential natural interannual variability. My research on P. falciparum malaria focuses on the analysis of the complexity behind the transmission dynamics of this multi-factorial disease. A deep understanding of such a complexity is possible through a holistic examination of the climatic, biological, socioeconomic, and demographic key-factors that are driving the fluctuations, changes and trends in malaria incidence. I propose a multi-model ensemble of malaria process-based models to offer useful information that could effectively guide decision-makers in risk assessment, malaria control investments and choice of interventions. I work on the integration of short-, medium- and long-term climate predictions into simulations of future changing scenarios, while helping to set up environment-informed systems at municipal level. My research thus provides a framework to: (a) compare the simulation outputs of several malaria process-based models with actual malaria morbidity profiles observed in several endemic- and epidemic-prone pilot sites in Colombia and Kenya; (b) explore the role that both climatic and non-climatic factors play in fluctuations and trends in malaria incidence, and analyze key confounders; (c) assess changing climate and future scenarios, and estimate the timing and possible magnitude of unexpected malaria outbreaks; (d) investigate current decision making processes, simulate the impact of indoor residual spraying campaigns, and provide quantitative goals for effective interventions; (e) conduct stability analysis; (f) pose and answer "what if" questions; and (g) stimulate an interactive learning environment to help decision makers learn.

When forced with increasing greenhouse gases, global climate models project changes to the seasonality of several key climate variables. These include delays in the phase of surface temperature, precipitation, and vertical motion indicating maxima and minima occurring later in the year. The changes also include an increase in the amplitude (or annual range) of low-latitude surface temperature and tropical precipitation and a decrease in the amplitude of high-latitude surface temperature and vertical motion. The aim of this thesis is to detail these changes, understand the links between them and ultimately relate them to simple physical mechanisms. At high latitudes, all of the global climate models of the CMIP3 intercomparison suite project a phase delay and amplitude decrease in surface temperature. Evidence is provided that the changes are mainly driven by sea ice loss: as sea ice melts during the 21st century, the previously unexposed open ocean increases the effective heat capacity of the surface layer, slowing and damping the temperature response at the surface. In the tropics and subtropics, changes in phase and amplitude are smaller and less spatially uniform than near the poles, but they are still prevalent in the models. These regions experience a small phase delay, but an amplitude increase of the surface temperature cycle, a combination that is inconsistent with changes to the effective heat capacity of the system. Evidence suggests that changes in the tropics and subtropics are linked to changes in surface heat fluxes. The next chapter investigates the nature of the projected phase delay and amplitude increase of precipitation using AGCM experiments forced by SST perturbations representing idealizations of the changes in annual mean, amplitude, and phase as simulated by CMIP5 models. A uniform SST warming is sufficient to force both an amplification and a delay of the annual cycle of precipitation. The amplification is due to an increase in the annual mean vertical water vapor gradient, while the delay is linked to a phase delay in the annual cycle of the circulation. A budget analysis of this simulation reveals a large degree of similarity with the CMIP5 results. In the second experiment, only the seasonal characteristics of SST are changed. For an amplified annual cycle of SST there is an amplified annual cycle of precipitation, while for a delayed SST there is a delayed annual cycle of precipitation. Assuming that SST changes can entirely explain the seasonal precipitation changes, the AGCM simulations suggest that the annual mean warming explains most of the amplitude increase and much of the phase delay in the CMIP5 models. However, imperfect agreement between the changes in the SST-forced AGCM simulations and the CMIP5 coupled simulations suggests that coupled effects may play a significant role. Finally, the connections between changes in the seasonality of precipitation, temperature and circulation are studied in the tropics using models of varying complexity. These models include coupled model simulations with idealized forcing, a simple, semi-empirical model to describe the effect of land-ocean interactions, an aquaplanet model, and a dry, dynamical model. Each gives insights into the projected CMIP changes. Taken together they suggest that changes in the amplitude of vertical motions are consistent with a weakening of the annual mean circulation and can explain part of the changes in the amplitude of precipitation over both ocean and land, when combined with the thermodynamic effect described previously. By increasing the amplitude of the annual cycle of surface winds, the changes in circulation may also increase the amplitude of the surface temperature via the surface energy balance. The delay in the phase of circulation directly leads to a delay in the phase of precipitation, especially over ocean.

Climatology is a paradigmatic complex systems science. Understanding the global climate involves tackling problems in physics, chemistry, economics, and many other disciplines. I argue that complex systems like the global climate are characterized by certain dynamical features that explain how those systems change over time. A complex system's dynamics are shaped by the interaction of many different components operating at many different temporal and spatial scales. Examining the multidisciplinary and holistic methods of climatology can help us better understand the nature of complex systems in general. Questions surrounding climate science can be divided into three rough categories: foundational, methodological, and evaluative questions. "How do we know that we can trust science?" is a paradigmatic foundational question (and a surprisingly difficult one to answer). Because the global climate is so complex, questions like "what makes a system complex?" also fall into this category. There are a number of existing definitions of `complexity,' and while all of them capture some aspects of what makes intuitively complex systems distinctive, none is entirely satisfactory. Most existing accounts of complexity have been developed to work with information-theoretic objects (signals, for instance) rather than the physical and social systems studied by scientists. Dynamical complexity, a concept articulated in detail in the first third of the dissertation, is designed to bridge the gap between the mathematics of contemporary complexity theory (in particular the formalism of "effective complexity" developed by Gell-Mann and Lloyd [2003]) and a more general account of the structure of science generally. Dynamical complexity provides a physical interpretation of the formal tools of mathematical complexity theory, and thus can be used as a framework for thinking about general problems in the philosophy of science, including theories, explanation, and lawhood. Methodological questions include questions about how climate science constructs its models, on what basis we trust those models, and how we might improve those models. In order to answer questions about climate modeling, it's important to understand what climate models look like and how they are constructed. Climate model families are significantly more diverse than are the model families of most other sciences (even sciences that study other complex systems). Existing climate models range from basic models that can be solved on paper to staggeringly complicated models that can only be analyzed using the most advanced supercomputers in the world. I introduce some of the central concepts in climatology by demonstrating how one of the most basic climate models might be constructed. I begin with the assumption that the Earth is a simple featureless blackbody which receives energy from the sun and releases it into space, and show how to model that assumption formally. I then gradually add other factors (e.g. albedo and the greenhouse effect) to the model, and show how each addition brings the model's prediction closer to agreement with observation. After constructing this basic model, I describe the so-called "complexity hierarchy" of the rest of climate models, and argue that the sense of "complexity" used in the climate modeling community is related to dynamical complexity. With a clear understanding of the basics of climate modeling in hand, I then argue that foundational issues discussed early in the dissertation suggest that computation plays an irrevocably central role in climate modeling. "Science by simulation" is essential given the complexity of the global climate, but features of the climate system--the presence of non-linearities, feedback loops, and chaotic dynamics--put principled limits on the effectiveness of computational models. This tension is at the root of the staggering pluralism of the climate model hierarchy, and suggests that such pluralism is here to stay, rather than an artifact of our ignorance. Rather than attempting to converge on a single "best fit" climate model, we ought to embrace the diversity of climate models, and view each as a specialized tool designed to predict and explain a rather narrow range of phenomena. Understanding the climate system as a whole requires examining a number of different models, and correlating their outputs. This is the most significant methodological challenge of climatology. Climatology's role contemporary political discourse raises an unusually high number of evaluative questions for a physical science. The two leading approaches to crafting policy surrounding climate change center on mitigation (i.e. stopping the changes from occurring) and adaptation (making post hoc changes to ameliorate the harm caused by those changes). Crafting an effective socio-political response to the threat of anthropogenic climate change, however, requires us to integrate multiple perspectives and values: the proper response will be just as diverse and pluralistic as the climate models themselves, and will incorporate aspects of both approaches. I conclude by offering some concrete recommendations about how to integrate this value pluralism into our socio-political decision making framework.

The hydroclimate of East Africa shows distinctive variabilities on seasonal to decadal time scales and poses a great challenge to climatologists attempting to project its response to anthropogenic emissions of greenhouse gases (GHGs). Increased frequency and intensity of droughts over East Africa in recent decades raise the question of whether the drying trend will continue into the future. To address this question, we first examine the decadal variability of the East African rainfall during March to May (MAM, the major rainy season in East Africa) and assess how well a series of models simulate the observed features. Observational results show that the drying trend during MAM is associated with decadal natural variability of sea surface temperature (SST) variations over the Pacific Ocean. The multimodel mean of the SST forced, Coupled Model Intercomparison Project Phase 5 (CMIP5) AMIP experiment models reproduces both the climatological annual cycle and the drying trend in recent decades. The fully coupled models from the CMIP5 historical experiment, however, have systematic errors in simulating the East African rainfall annual cycle by underestimating the MAM rainfall while overestimating the October to December (OND, the second rainy season in East Africa) rainfall. The multimodel mean of the historical coupled runs of the MAM rainfall anomalies, which is the best estimate of the radiatively forced change, shows a weak wetting trend associated with anthropogenic forcing. However, the SST anomaly pattern associated with the MAM rainfall has large discrepancies with the observations. The errors in simulating the East African hydroclimate with coupled models raise questions about how reliable model projections of future East African climate are. This motivates a fundamental study of why East African climate is the way it is and why coupled models get it wrong. East African hydroclimate is characterized by a dry annual mean climatology compared to other deep tropical land areas and a bimodal annual cycle with the major rainy season during MAM (often called the ``long rains'' by local people) and the second during OND (the ``short rains''). To explore these distinctive features, we use the ERA Interim Re Analysis data to analyze the associated annual cycles of atmospheric convective stability, circulation and moisture budget. The atmosphere over East Africa is found to be convectively stable, in general, year round but with an annual cycle dominated by the surface moist static energy (MSE), which is in phase with the precipitation annual cycle. Throughout the year, the atmospheric circulation is dominated by a pattern of convergence near the surface, divergence in the lower troposphere and convergence again at upper levels. Consistently, the convergence of the vertically integrated moisture flux is mostly negative across the year, but becomes weakly positive in the two rainy seasons. It is suggested the semi-arid/arid climate in East Africa and its bimodal rainfall annual cycle can be explained by the ventilation mechanism, in which the atmospheric convective stability over East Africa is controlled by the import of low MSE air from the relatively cool Indian Ocean off the coast and the cold winter hemisphere. During the rainy seasons, however, the off coast SST increases (and is warmest during the long rains season) and the northerly or southerly weakens, and consequently the air imported into East Africa becomes less stable. The MSE framework is then applied to study the coupling induced bias of the East African rainfall annual cycle often found in CMIP3/5 coupled models that overestimates the OND rainfall and underestimates the MAM rainfall, by comparing the historical (coupled) and the AMIP runs (SST forced) for each model. It is found that a warm north and cold south SST bias over the Indian Ocean induced in coupled models is responsible for the dry MAM rainfall bias over East Africa while the ocean dynamics induced warm west and cold east SST bias over the Indian Ocean contributes to the wet OND rainfall bias in East Africa. Finally, to understand the East African regional climate in the context of the broader tropical climate and circulation, zonal momentum balance of the tropical atmospheric circulation during the global monsoon mature months (January and July) are analyzed in three dimensions based on the ERA-Interim Re-Analysis. It is found that the dominant terms in the balance of the atmospheric boundary layer (ABL) in both months are the pressure gradient force, the Coriolis force and friction. The nonlinear advection term plays a significant role only in the Asian summer monsoon regions including off East Africa. In the upper troposphere, the pressure gradient force, the Coriolis force and nonlinear advection are the dominant terms. The transient eddy force and the residual force (which can be explained as convective momentum transfer over open oceans) are secondary yet can not be neglected near the equator. Zonal mean equatorial upper troposphere easterlies are maintained by the absolute angular momentum advection associated with the cross equatorial Hadley circulation. Equatorial upper troposphere easterlies over the Asian monsoon regions are also controlled by the absolute angular momentum advection but are mainly maintained by the pressure gradient force in January. The equivalent linear Rayleigh friction, which is widely applied in simple tropical models, is calculated and the corresponding spatial distribution of local coefficient and damping time scale are estimated from the linear regression. It is found that the linear momentum model is in general capable of crudely describing the tropical atmospheric circulation dynamics yet the caveat should be kept in mind that the friction coefficient is not uniformly distributed and is even negative in some regions.