In this study, the space-time relationship of precipitation fields is examined by testing the Taylor's "frozen field" hypothesis (TH). Specifically, the hypothesis supposes that if a spatio-temporal precipitation field with a stationary covariance Cov(r,tau) in both space r and time tau, moves with a constant velocity v, then the temporal covariance at time lag tau is equal to the spatial covariance at space lag v tau, that is, Cov(0;tau) = Cov(v tau, 0). Of specific interest is whether there is a cut-off or decorrelation time scale for which the TH holds for a given mean flow velocity v. The validity of the TH is tested for precipitation fields using high-resolution gridded NEXRAD radar reflectivity data over southeastern United States by employing two different statistical approaches. The first method is based upon rigorous hypothesis testing while the second is based on a simple correlation analysis, which neglects possible dependencies in the correlation estimates. The data-set has an approximate horizontal resolution of 4 km x 4 km and a temporal resolution of 15 minutes, while the time period of study is 4 days. The results of both statistical methods suggest that the TH might hold for the shortest space and time scales resolved by the data (4 km and 15 minutes), but that it does not hold for longer periods or larger spatial scales. The fidelity of global climate models in accurately simulating seasonal mean precipitation in the tropics is investigated by comparisons with satellite observations. Specifically, six-year long (2000-2005) simulations are performed using a high-resolution (36-km) Weather Research Forecast (WRF) model and the Community Atmosphere Model (CAM) at T85 spatial resolution and the results are compared with satellite observations from the Tropical Rainfall Measuring Mission (TRMM). The primary goal is to study the annual cycle of rainfall over four land regions of the tropics namely, the Indian monsoon, the Amazon, tropical Africa and the North American monsoon. The results indicate that the WRF model systematically underestimates the magnitude of monthly mean rainfall over most Tropical land regions but gets the seasonal timing right. On the other hand, CAM produces rainfall magnitudes that are closer to the observations but the rainfall peak leads or lags the observations by a month or two. Some of these regional biases can be attributed to erroneous circulation and moisture surpluses/deficits in the lower troposphere in both models. Overall, the results seem to indicate that employing a higher spatial resolution (36 km) does not significantly improve simulation of precipitation. We speculate that a combination of several physics parameterizations and lack of model tuning gives rise to the observed differences between the models and the observations.

Contemporary observations show an increase in the melting of the Greenland Ice Sheet (GrIS) since the early 21st century. Located near the critical sites of oceanic deep convection and deep water formation, the melting of the GrIS has the potential to directly impact the Atlantic Meridional Overturning Circulation (AMOC) by freshening ocean surface waters in these regions. The majority of the Coupled Model Intercomparison Project Phase 5 (CMIP5) models project a decline in AMOC strength by 10-50% during the 21st century, in response to the increase in atmospheric greenhouse gas (GHG) concentrations. However, due to the simple treatment of polar ice sheets and the lack of a dynamical ice sheet component in these models, these projections likely underestimated the impacts of the GrIS melt, leading to uncertainty in projecting future AMOC evolution and climate change around Greenland. To better understand the impact of the GrIS melt on the AMOC, we perform a series of 21st century projection runs with a state-of-the-art Earth System Model-GFDL ESM2Mb. We consider a medium and a high Representative Concentration Pathway (RCP) scenario (RCP4.5 and RCP8.5, respectively). Unlike the CMIP5-standard RCP runs which included only radiative forcing, the new model experiments are also forced with additional and potentially more realistic meltwater discharge from the GrIS. This meltwater discharge is estimated based on a model-based relationship between the GrIS surface melt and the 500hPa atmospheric temperature anomalies over Greenland. The model simulations indicate that compared to the RCP4.5-only and RCP8.5-only projections, the additional melt water from the GrIS can further weaken the AMOC, but with a relatively small magnitude. The reason is that radiative forcing already weakens the deep convection and deep water formation in the North Atlantic, therefore limiting the magnitude of further weakening of AMOC due to the additional meltwater. The modeling results suggest that the AMOC's sensitivity to freshwater forcing due to the GrIS melt is highly dependent on the location and strength of oceanic deep convection sites in ESM2Mb as well as the pathways of the meltwater towards these regions. The additional meltwater contributes to the minimum surface warming (so-called "warming hole") south of Greenland. These simulations with ESM2Mb contribute to the Atlantic Meridional Overturning Circulation Model Intercomparison Project (AMOCMIP), a community effort between international modeling centers to investigate the impacts of the melting of the GrIS on the AMOC and quantify the associated uncertainty.

The long warming transition from the Last Ice Age into the present Interglacial period, the last deglaciation, holds the key to our understanding of future abrupt climate change. In the last decades, a great effort has been put into deciphering the linkage between freshwater fluxes from melting ice sheets and rapid shifts in global ocean-atmospheric circulation that characterized this puzzling climate period. In particular, the regional expressions of climate change in response to freshwater forcing are still largely unresolved. This projects aims at evaluating the environmental, hydro-climatic and oceanographic response in the Eastern North Atlantic domain to freshwater fluxes from the Scandinavian Ice Sheet during the last deglaciation (~19,000-11,000 years ago). The results presented in this thesis involve an overview of the regional representations of climate change across rapid climatic transitions and provide the groundwork to better understand spatial and temporal propagations of past atmospheric and ocean perturbations. Specifically, this thesis comprises i) a comparison of pollenstratigraphic records from densely 14C dated lake sediment sequences, which provides insight into the regional sensitivity of North European vegetation to freshwater forcing in the Nordic Seas around the onset of the Younger Dryas stadial (~12,900 years ago); ii) a reconstruction of North European hydro-climate, which, together with transient climate simulations, shed light on the mechanisms and regionality of climate shortly prior to the transition into the Younger Dryas stadial; iii) studies of a ~1250-year long glacial varve chronology, which provides an accurate timing for the sudden drainage of proglacial freshwater stored in the former ice-dammed Baltic Ice Lake into the North Atlantic Ocean; iv) a 5000-year long terrestrial-marine reconstruction of Eastern North Atlantic hydro-climate and oceanographic changes that clarifies the hitherto elusive relationship between freshwater forcing and the transient behaviour of the North Atlantic overturning circulation system. The results presented in this thesis provide new important temporal constraints on the events that punctuated the last deglaciation in Northern Europe, and give a clearer understanding of the ocean – atmosphere – ice-sheet feedbacks that were at work in the North Atlantic. This increases our understanding of how the Earth climate system functions in more extreme situations. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 3: In press. Paper 4: Manuscript.</p>

The Wind-Evaporation-Sea Surface Temperature (WES) feedback is believedto play an important role in the tropics, where climate variability is governed byatmosphere-ocean coupled interactions. This dissertation reports on studies to distinctlyisolate the WES feedback mechanism over tropical oceans using a modiedversion of an NCAR-Community Climate Model (CCM3) thermodynamically coupledto a slab ocean model, where the WES feedback is deliberately suppressed inthe bulk aerodynamic formulation for surface heat uxes. A comparison of coupledintegrations using the modified WES-off CCM3 to those carried out using the standardCCM3 conclusively identifies the role of the WES feedback in enhancing theinter-annual variability over deep tropical oceans and the westward propagation ofthe equatorial annual cycle. An important role for near surface humidity in tropicalclimate variability in enhancing inter-annual variability and in sustaining the equatorialannual cycle is also suggested. Statistical analyses over the tropical Atlanticreveal that the free coupled meridional mode of the Atlantic Ocean is amplified in thepresence of the WES feedback. Similar analyses of coupled model integrations, whenforced with an articial El Ni~no Southern Oscillation (ENSO)-like SST cycle in tropicalPacific, reveal that only in the presence of the WES feedback is the meridionalmode the preferred mode of response of the Atlantic to ENSO forcings. It is also foundthat WES feedback reinforces the tendency of the ITCZ to stay north of the equator over the Atlantic during El-Nino events. Comparative studies between Last GlacialMaximum (LGM) equivalent imposed northern hemispheric sea-ice experiments withthe WES-off model and the standard model indicate a dominant role for the WESfeedback in the southward shift of the ITCZ as indicated by paleo-climate records.However, it is found not to be the sole thermodynamic mechanism responsible for thepropagation of high latitude cold SST anomalies to the tropics, suggesting significantroles for other mechanisms in the tropical response to high latitude changes.

The Wind-Evaporation-Sea Surface Temperature (WES) feedback is believedto play an important role in the tropics, where climate variability is governed byatmosphere-ocean coupled interactions. This dissertation reports on studies to distinctlyisolate the WES feedback mechanism over tropical oceans using a modiedversion of an NCAR-Community Climate Model (CCM3) thermodynamically coupledto a slab ocean model, where the WES feedback is deliberately suppressed inthe bulk aerodynamic formulation for surface heat uxes. A comparison of coupledintegrations using the modified WES-off CCM3 to those carried out using the standardCCM3 conclusively identifies the role of the WES feedback in enhancing theinter-annual variability over deep tropical oceans and the westward propagation ofthe equatorial annual cycle. An important role for near surface humidity in tropicalclimate variability in enhancing inter-annual variability and in sustaining the equatorialannual cycle is also suggested. Statistical analyses over the tropical Atlanticreveal that the free coupled meridional mode of the Atlantic Ocean is amplified in thepresence of the WES feedback. Similar analyses of coupled model integrations, whenforced with an articial El Ni~no Southern Oscillation (ENSO)-like SST cycle in tropicalPacific, reveal that only in the presence of the WES feedback is the meridionalmode the preferred mode of response of the Atlantic to ENSO forcings. It is also foundthat WES feedback reinforces the tendency of the ITCZ to stay north of the equator over the Atlantic during El-Nino events. Comparative studies between Last GlacialMaximum (LGM) equivalent imposed northern hemispheric sea-ice experiments withthe WES-off model and the standard model indicate a dominant role for the WESfeedback in the southward shift of the ITCZ as indicated by paleo-climate records.However, it is found not to be the sole thermodynamic mechanism responsible for thepropagation of high latitude cold SST anomalies to the tropics, suggesting significantroles for other mechanisms in the tropical response to high latitude changes.

Due to factors such as spatial discretization and the parameterization of certain processes, the presence of bias in models of the Earth's atmosphere is unavoidable. Whether we are selecting a model to explain past phenomenon, forecast weather patterns, or make inferences about the future, the target of any selection process is to minimize the discrepancies between model output and observations. Some discrepancies have a greater effect on the scatter of model predictions though. We exemplify this in the case of CO2 forced warming using multivariate empirical orthogonal functions (EOF), created using an ensemble of plausible parameter configurations of CAM3.1. When subjecting this ensemble to a doubling of atmospheric CO2, some EOFs exhibit significantly higher correlation than others with the resulting increase in mean global surface temperature. Therefore, there are discernible bias patterns that effect its predictive scatter. By targeting these patterns in the model evaluation process, it is plausible to use this information to constrain the resulting range of predictions. We take a first step towards showing this by creating a metric to evaluate model skill based on these EOFs and their correlation to a model's sensitivity to CO2 forcing. Using model output, for which we know the resulting temperature increase, as a surrogate for observations in this metric, the resulting distribution of skill scores indeed agreement in sensitivity to CO2 forcing. / text

Land surface covers only 30% of the global surface, but contributes largely to the intricacy of the climate system by exchanging water and energy with the overlying atmosphere. The partitioning of incident solar radiation among various components at the land surface, especially vegetation and underlying soil, determines the energy absorbed by vegetation, evapotranspiration, partitioning between surface sensible and latent heat fluxes, and the energy and water exchange between the land surface and the atmosphere. Because of its significance in climate model, land surface model solar radiation partitioning scheme should be evaluated in order to ensure its accuracy in reproducing these naturally complicated processes. However, few studies evaluated this part of climate model. This study examines a land surface solar radiation partitioning scheme, i.e., that of the Community Land Model version 4 (CLM4) with coupled carbon and nitrogen cycles. Taking advantage of multiple remote sensing fraction of absorbed photosynthetically active radiation (FPAR) datasets, ground observations and a unique 28-year FPAR dataset derived from the Global Inventory Modeling and Mapping Studies (GIMMS) Normalized Difference Vegetation Index (NDVI) dataset, we evaluated the CLM4 FPAR’s seasonal cycle, diurnal cycle, long-term trends and spatial patterns. Our findings show the model roughly agrees with observations in the seasonal cycle , long-tern trend and spatial patterns but does not reproduce the diurnal cycle. Discrepancies also exist in seasonality magnitudes, peak value months and spatial heterogeneity. We identified the discrepancy in the diurnal cycle as due to the absence of dependence on sun angle in the model. Implementation of sun angle dependence in a one-dimensional (1-D) model is proposed. The need for better relating vegetation to climate in the model indicated by long-term trends is also noted. Evaluation of the CLM4 land surface solar radiation partitioning scheme using remote sensing and site level FPAR datasets provides targets for future development in its representation of this naturally complicated process. / text

A commonly noted problem in the simulation of warm season convection in the North American monsoon region has been the inability of atmospheric models at the meso- scales (10 s to 100 s of kilometers) to simulate organized convection, principally mesoscale convective systems. With the use of convective parameterization, high precipitation biases in model simulations are typically observed over the peaks of mountain ranges. To address this issue, the Kain-Fritsch (KF) cumulus parameterization scheme has been modified with new diagnostic equations to compute the updraft velocity, the convective available potential energy closure assumption, and the convective trigger function. The scheme has been adapted for use in the Weather Research and Forecasting (WRF). A numerical weather prediction-type simulation is conducted for the North American Monsoon Experiment Intensive Observing Period 2 and a regional climate simulation is performed, by dynamically downscaling. In both of these applications, there are notable improvements in the WRF model-simulated precipitation due to the better representation of organized, propagating convection. The use of the modified KF scheme for atmospheric model simulations may provide a more computationally economical alternative to improve the representation of organized convection, as compared to convective-permitting simulations at the kilometer scale or a super-parameterization approach.

Here we analyze climate variability using instrumental, paleoclimate (proxy), and the latest climate model data to understand more about the sources and impacts of low-frequency climate variability. Understanding the drivers of climate variability at interannual to century timescales is important for studies of climate change, including analyses of detection and attribution of climate change impacts. Additionally, correctly modeling the sources and impacts of variability is key to the simulation of abrupt change (Alley et al., 2003) and extended drought (Seager et al., 2005; Pelletier and Turcotte, 1997; Ault et al., 2014). In Appendix A, we employ an Earth system model (GFDL-ESM2M) simulation to study the impacts of a weakening of the Atlantic meridional overturning circulation (AMOC) on the climate of the American Tropics. The AMOC drives some degree of local and global internal low-frequency climate variability (Manabe and Stouffer, 1995; Thornalley et al., 2009) and helps control the position of the tropical rainfall belt (Zhang and Delworth, 2005). We find that a major weakening of the AMOC can cause large-scale temperature, precipitation, and carbon storage changes in Central and South America. Our results suggest that possible future changes in AMOC strength alone will not be sufficient to drive a large-scale dieback of the Amazonian forest, but this key natural ecosystem is sensitive to dry-season length and timing of rainfall (Parsons et al., 2014). In Appendix B, we compare a paleoclimate record of precipitation variability in the Peruvian Amazon to climate model precipitation variability. The paleoclimate (Lake Limón) record indicates that precipitation variability in western Amazonia is ‘red’ (i.e., increasing variability with timescale). By contrast, most state-of-the-art climate models indicate precipitation variability in this region is nearly'‘white' (i.e., equally variability across timescales). This paleo-model disagreement in the overall structure of the variance spectrum has important consequences for the probability of multi-year drought. Our lake record suggests there is a significant background threat of multi-year, and even decade-length, drought in western Amazonia, whereas climate model simulations indicate most droughts likely last no longer than one to three years. These findings suggest climate models may underestimate the future risk of extended drought in this important region. In Appendix C, we expand our analysis of climate variability beyond South America. We use observations, well-constrained tropical paleoclimate, and Earth system model data to examine the overall shape of the climate spectrum across interannual to century frequencies. We find a general agreement among observations and models that temperature variability increases with timescale across most of the globe outside the tropics. However, as compared to paleoclimate records, climate models generate too little low-frequency variability in the tropics (e.g., Laepple and Huybers, 2014). When we compare the shape of the simulated climate spectrum to the spectrum of a simple autoregressive process, we find much of the modeled surface temperature variability in the tropics could be explained by ocean smoothing of weather noise. Importantly, modeled precipitation tends to be similar to white noise across much of the globe. By contrast, paleoclimate records of various types from around the globe indicate that both temperature and precipitation variability should experience much more low-frequency variability than a simple autoregressive or white-noise process. In summary, state-of-the-art climate models generate some degree of dynamically driven low-frequency climate variability, especially at high latitudes. However, the latest climate models, observations, and paleoclimate data provide us with drastically different pictures of the background climate system and its associated risks. This research has important consequences for improving how we simulate climate extremes as we enter a warmer (and often drier) world in the coming centuries; if climate models underestimate low-frequency variability, we will underestimate the risk of future abrupt change and extreme events, such as megadroughts.

An ensemble of 11 dynamically downscaled CMIP3 GCMs under A2 projection scenario are first bias corrected for the historic (1971-2000) and scenario (2041-2070) period using a Scaled Distribution Mapping (SDM) technique, that preserves the relative change in the monthly mean and variance of precipitation and any model trends in temperature to generate an ensemble of streamflow projections across 3 catchments in the Colorado River basin - Upper Colorado at Lees Ferry, Salt and Verde. The hydroclimatic projections obtained from this method are compared against an existing ensemble of 15 Bias Corrected and Spatially Disaggregated (BCSD) CMIP3 models under A2 projection scenario developed by the Bureau of Reclamation (BOR). The confidence in the DD Ens. stems from its ability to represent historical flow quantiles better than BCSD Ens. Across all three basins, the mean of the dynamically downscaled ensemble (DD Ens.) projects a decrease in both monsoon and winter projected precipitation as compared to mean of the statistically downscaled ensemble (BCSD Ens.). For the Upper Colorado, both Ens. show a shift in peak hydrograph from June to May due to earlier snowmelt, but a projected decrease in precipitation (-5%) by DD Ens. as compared to a slight increase (+2%) by BCSD Ens. results in a lower April snow water equivalent (SWE) and reduced streamflows (14% by DD Ens. as compared to 5% by BCSD Ens.). The streamflow decrease over the Upper Colorado River basin, quantified by both the mean and the spread of the ensemble. is representative in high flows and flows during moist conditions. For smaller basins like Salt and Verde, DD Ens. shows a greater decrease (-11%) in precipitation than BCSD Ens. (-2%), which results in lower peak hydrograph during March and significantly reduced streamflows (-20%&-14% for Salt and Verde by DD Ens. as compared to -3% by BCSD Ens.). This decrease is more substantial in high flows, but occurs across all streamflow quantiles. The future streamflow projection, quantified by the spread of the DD Ens. presents the shifting of the streamflow range downward to be drier in the future.