AGU 2013 Fall Meeting Abstracts

Columbia Water Center Research Scientists to Present at the

AGU Fall Meeting 2013 | San Francisco | December 9-13

The American Geophysical Union fall conference will be taking place in San Francisco California from December 9th – 13th, 2013.  Several research scientists from the Columbia Water Center will be attending and presenting their work.  Some of the topics that will be covered include extreme weather patterns, floods and their risk factors, agricultural water management in India, climate change mitigation strategies for reservoirs and dams, nitrogen scarcity in sub-Sahara Africa and new methods of estimating trends in evapotranspiration, just to name a few.  Below is a full list of the abstract submissions, their authors and affiliated institutions along with the dates and times for each session.

Monday December 9, 2013

NG11A. Statistical Geophysics: Natural Hazards and Fracking I Posters [SWIRL_CM.CU]
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“The Spatial Scaling of Global Rainfall Extremes”
ID: 1815669

AUTHORS/INSTITUTIONS: N. Devineni, U. Lall, B. Rahill-Marier, Columbia Univ-Water Ctr, New York, New York, UNITED STATES; C. Xi, Hydrology, Hohai University, Nanjing, CHINA

BODY: Floods associated with severe storms are a significant source of risk for property, life and supply chains. These property losses tend to be determined as much by the duration of flooding as by the depth and velocity of inundation. High duration floods are typically induced by persistent rainfall (upto 30 day duration) as seen recently in Thailand, Pakistan, the Ohio and the Mississippi Rivers, France, and Germany. Events related to persistent and recurrent rainfall appear to correspond to the persistence of specific global climate patterns that may be identifiable from global, historical data fields, and also from climate models that project future conditions. A clear understanding of the space-time rainfall patterns for events or for a season will enable in assessing the spatial distribution of areas likely to have a high/low inundation potential for each type of rainfall forcing. In this paper, we investigate the statistical properties of the spatial manifestation of the rainfall exceedances. We also investigate the connection of persistent rainfall events at different latitudinal bands to large-scale climate phenomena such as ENSO. Finally, we present the scaling phenomena of contiguous flooded areas as a result of large scale organization of long duration rainfall events. This can be used for spatially distributed flood risk assessment conditional on a particular rainfall scenario. Statistical models for spatio-temporal loss simulation including model uncertainty to support regional and portfolio analysis can be developed.

GC11A. Paleoclimate, Observations, and Models: Water Resource Management Under Climate Variability and Change I Posters
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“Diagnostics of Interannual-to-Interdecadal Climate and Streamflow Variability: Applications to Reservoir Management over NW India”
ID: 1798047

AUTHORS/INSTITUTIONS: A.W. Robertson, A.M. Greene, IRI, Columbia University, Palisades, New York, UNITED STATES;

E.R. Cook, Tree Ring Laboratory, Columbia University, Palisades, New York, UNITED STATES;
U. Lall, M. Lu, Earth & Env Eng, Columbia University, New York, New York, UNITED STATES;
M. Ghil, D.A. Kondrashov, Atmospheric & Ocean Sciences, UCLA, Los Angeles, California, UNITED STATES

BODY: Multi-year storage reservoirs must be managed in the face of weather and climate variability across time scales ranging from daily weather to interannual climate. While seasonal climate may contain a predictable component associated with the El Nino-Southern Oscillation (ENSO), longer time scales are not yet usefully predictable, nor is the interannual-to-interdecadal power spectrum well estimated from observed data. In addition, climate simulations from general circulation models (GCMs) are often lacking in their ability to generate realistic hydroclimate variability across time scales, especially at small spatial scales. These issues are critical for climate change adaptation planning in water management, where realistic estimates of climate and stream flow variability are required.

For the Bhakra reservoir in NW India, we develop estimates of climate and stream flow variability, including the interannual-to-interdecadal power spectrum, based on (1) instrumental stream flow records of the Sutlej river, 1963–2010; (2) tree ring reconstructions of the Sutlej flow back to 1321; and (3) multi-century control simulations of precipitation-minus-evaporation made with several coupled ocean-atmosphere GCMs archived in the IPCC CMIP5 database. By comparing these observed, paleo-proxy, and GCM-based estimates, we shed light on the ability of GCMs to simulate realistic hydroclimate variability over the Indus basin, as well as on the nature of tree-ring based streamflow reconstructions. In addition to these estimates of the variability spectrum, we explore the use of a nonlinear, multi-level stochastic polynomial inverse model to bridge between these different datasets.

H11A. Budyko Hypothesis Revisited: Advances, Challenges and Opportunities Posters8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“The Budyko and complementary relationships in the large-scale coupled land-atmosphere system”
ID: 1803090

AUTHORS/INSTITUTIONS: B.R. Lintner, Environmental Sciences, Rutgers, New Brunswick, New Jersey, UNITED STATES;

P. Gentine, Earth and Environmental Engineering, Columbia University, New York, New York, UNITED STATES;

K.L. Findell, Geophysical Fluid Dynamics Laboratory, Princeton, New Jersey, UNITED STATES;
G. Salvucci, Earth and Environment, Boston University, Boston, Massachusetts, UNITED STATES;
P. D’Odorico, Environmental Sciences, University of Virginia, Charlottesville, Virginia, UNITED STATES;
G. Sivandran, Civil, Environmental, and Geodetic Engineering, Ohio State University, Columbus, Ohio, UNITED STATES

BODY: The Budyko and complementary relationships reflect emergent behaviors in the hydrologic cycle. Although these relationships appear to hold across a wide range of climate and ecosystem conditions, they remain largely unexplained. Here we use an idealized prototype of large-scale land-atmosphere coupling developed in prior work to derive (at an implicitly large and climatic scales) expressions analogous to the Budyko and complementary relationships. In contrast to many prior analytic approaches, the approach here considers convection and precipitation implicitly as part of the solution rather than as an external forcing. The analytic solutions permit diagnosis of the sensitivity of the Budyko and complementary relationships to various atmospheric and land surface processes. In particular, we use these relationships to investigate how anthropogenic influences, including large-scale irrigation and global warming, may be expected to impact the hydrologic cycle. Moreover, using simplified ecohydrological modeling, we show how vegetation adaptation further constrains the stability of the Budyko curve. Vegetation effectively acts as a seasonal low-pass filter, producing in-phase energy (radiation) and water (precipitation) availability that in turn maximize carbon gain.

H11L. Advances in Understanding Land-Atmosphere Interactions I [SWIRL_GS]
8:00 AM – 10:00 AM; 3022 (Moscone West)

“Unique relation between surface-limited evaporation and relative humidity profiles holds in both field data and climate model simulations”
ID: 1819277

AUTHORS/INSTITUTIONS: G. Salvucci, A.J. Rigden, Earth and Environment, Boston Univ, Boston, Massachusetts, UNITED STATES;

P. Gentine, Earth and Environmental Engineering, Columbia, New York, New York, UNITED STATES;
B.R. Lintner, Environmental Sciences, Rutgers, New Brunswick, New Jersey, UNITED STATES

BODY: A new method was recently proposed for estimating evapotranspiration (ET) from weather station data without requiring measurements of surface limiting factors (e.g. soil moisture, leaf area, canopy conductance) [Salvucci and Gentine, 2013, PNAS, 110(16): 6287–6291]. Required measurements include diurnal air temperature, specific humidity, wind speed, net shortwave radiation, and either measured or estimated incoming longwave radiation and ground heat flux. The approach is built around the idea that the key, rate-limiting, parameter of typical ET models, the land-surface resistance to water vapor transport, can be estimated from an emergent relationship between the diurnal cycle of the relative humidity profile and ET. The emergent relation is that the vertical variance of the relative humidity profile is less than what would occur for increased or decreased evaporation rates, suggesting that land-atmosphere feedback processes minimize this variance. This relation was found to hold over a wide range of climate conditions (arid to humid) and limiting factors (soil moisture, leaf area, energy) at a set of Ameriflux field sites.

H11L. Advances in Understanding Land-Atmosphere Interactions I [SWIRL_GS]
8:00 AM – 10:00 AM; 3022 (Moscone West)

“Impact of soil moisture-atmosphere interactions on surface temperature distribution”
ID: 1819492

AUTHORS/INSTITUTIONS: A.M. Berg, B.R. Lintner, Rutgers University, New Brunswick, New Jersey, UNITED STATES;

A.M. Berg, K.L. Findell, GFDL, Princeton, New Jersey, UNITED STATES;
P. Gentine, Columbia University, New York, New York, UNITED STATES;
S. Malyshev, Princeton University, Princeton, New Jersey, UNITED STATES

BODY: Land-atmosphere interactions are a key physical process in the climate system. One of the critical variables involved in these interactions is soil moisture, as it partly controls radiative and turbulent heat fluxes to the atmosphere. Through these processes, soil moisture variability has the potential to feed back on near-surface hydroclimate, in particular temperature.

This study investigates simulations performed at GFDL in the frame of the GLACE-CMIP5 project to investigate soil moisture feedbacks : a coupled land-atmosphere model was run over 1950-2100 with transient forcings, prescribed SSTs (derived from the corresponding historical and future coupled CMIP5 simulations) and with either interactive soil moisture, or soil moisture prescribed to its 1971-2000 climatology. Here we compare in particular the two simulations over 1971-2000, isolating the effect of soil moisture dynamics (since soil moisture climatology is identical) on the simulated climate. We place the emphasis on the distribution of daily near-surface temperatures, and how the associated probability distribution function is shaped by soil moisture-atmosphere interactions.

We show that soil moisture dynamics strongly enhance both temperature mean and variability over apparent regional ’hotspots’. Moreover, higher-order distribution moments such as skewness and kurtosis are also significantly impacted: in particular, skewness generally becomes more positive with interactive soil moisture, suggesting an asymmetric impact on hot and cold anomalies. We interpret these changes by considering changes in the distributions of the surface radiative and turbulent fluxes. Importantly, the different temperature pdf parameters are not all affected at the same time or in a similar way in different regions. These different behaviors underscore the importance of analyzing all distribution moments to fully characterize the impacts of soil moisture-atmosphere interactions on surface temperature. In addition, we show that soil moisture dynamics impact daily temperature variability at different time over different regions in the model. The impacts of soil moisture dynamics on the distribution of surface temperatures have implications for the analysis (attribution, projections) of extreme temperature events.

Tuesday December 10, 2013

H21I. Uncertainty in Water Management, Part 2: Risk Analysis, Decision Support and Law, With Special Focus on Hydrometeorological Scaling From Continents to Watersheds II Posters [SWIRL_CU]
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“Understanding Scale in Flooding: The role of drainage area, heavy precipitation, and model resolution”
ID: 1810521

AUTHORS/INSTITUTIONS: T. Troy, N. Devineni, U. Lall, Columbia Water Center, The Earth Institute at Columbia University, New York, New York, UNITED STATES;

T. Troy, Civil and Environmental Engineering, Lehigh University, Bethlehem, Pennsylvania, UNITED STATES;

U. Lall, Earth and Environmental Engineering, Columbia University, New York, New York, UNITED STATES

BODY: Floods are costly natural disasters, causing significant loss of life and destruction of property. Understanding how floods will change with climate in the coming decades is therefore of considerable interest. Existing studies largely focus on changes in precipitation extremes and assume that this extends to floods. However, it is unclear that an increase in more intense precipitation automatically translates into an increase in flooding. For example, large-scale floods are not the result of extreme 1-day rainfall. Rather it is the long-term precipitation dynamics combined with basin characteristics that dictate the relationship between heavy rainfall and flood events. This study tackles two open questions that need to be explored to understand the appropriate approach to projections of future flooding.

First, we must understand how basin size influences the relationship between basin rainfall and the resulting floods. To tackle this question, we use historical precipitation and streamflow records for 165 basins in the Ohio River. Results indicate that 1-day rainfall and peak streamflow are highly correlated in small basins. For larger basins (>2000 square miles), 7-day rainfall has the strongest correlation with peak flood flows. This informs what variables must be considered when examining how changes in precipitation extremes may translate into flood events.

Second, many studies analyzing the response of hydrologic systems to climate change involve downscaling climate information and feeding it through a hydrologic model. This raises questions of the appropriate resolution for accurate flood modeling. In order to understand the effect spatial and temporal resolution have on modeled flooding, we performed model experiments with the Variable Infiltration Capacity (VIC) model. These experimental results can inform future studies that seek to understand how flood regimes will change with climate.

GC21A. Understanding and Quantifying Changes in the Water Cycle I Posters
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“Multi-decadal Estimation of Trends in Evapotranspiration from Weather Station Data using a New Approach”
ID: 1807158

AUTHORS/INSTITUTIONS: A.J. Rigden, G. Salvucci, Earth and Environment, Boston University, Boston, Massachusetts, UNITED STATES;

P. Gentine, Earth and Environmental Engineering, Columbia University, New York, New York, UNITED STATES

BODY: We apply a new method of estimating evapotranspiration using historical weather station data and analyze the associated trends in evapotranspiration in the context of climate change. The method used to estimate ET is described and validated at field sites elsewhere (Salvucci and Gentine (2013), PNAS, 110(16): 6287–6291). The approach is built around the idea that the key, rate-limiting, parameter of typical ET models, the land-surface resistance to water vapor transport, can be estimated from an emergent relationship between the diurnal cycle of the relative humidity profile and ET. The emergent relation is that the vertical variance of the relative humidity profile is less than what would occur for increased or decreased evaporation rates, suggesting that land-atmosphere feedback processes minimize this variance. This relation was found to hold over a wide range of climate conditions (arid to humid) and limiting factors (soil moisture, leaf area, energy). Using this relation, daily estimates of ET are obtained across the United States for the second half of the twentieth century using meteorological data measured at common weather stations, without requiring measurements of surface limiting factors (soil moisture, leaf area, canopy conductance). Required measurements include diurnal air temperature, specific humidity, wind speed, and net shortwave radiation. Using relatively simple models for the less commonly measured radiation terms (incoming long wave radiation, dependent on screen height air temperature and humidity, and ground heat flux, dependent on surface temperature), estimates of daily ET are made and compared with a water budget estimate of ET using UNH GRDC runoff dataset across the United States. The estimated ET trends (both annual and seasonal) are regional and the variability of the ET trends can be attributed to three terms: radiation (longwave down + net shortwave), surface resistance, and atmospheric resistance. An analysis of ET trends and the associated drivers of these trends will be presented.

H21N. Hydrological Change and Water Systems: Feedbacks, Prediction, and Experimental Management I (Virtual Option)
8:00 AM – 10:00 AM; 3002 (Moscone West)

“From Hydroclimatic Prediction to Negotiated and Risk Managed Water Allocation and Reservoir Operation”
ID: 1807893

AUTHORS/INSTITUTIONS: U. Lall, Earth & Env Eng, Columbia Univ, New York, New York, UNITED STATES

BODY: The availability of long lead climate forecasts that can in turn inform streamflow, agricultural, ecological and municipal/industrial and energy demands provides an opportunity for innovations in water resources management that go beyond the current practices and paradigms.

In a practical setting, managers seek to meet registered demands as well as they can. Pricing mechanisms to manage demand are rarely invoked. Drought restrictions and operations are implemented as needed, and pressures from special interest groups are sometimes accommodated through a variety of processes. In the academic literature, there is a notion that demand curves for different sectors could be established and used for “optimal management”. However, the few attempts to implement such ideas have invariably failed as elicitation of demand elasticity and socio-political factors is imperfect at best.

In this talk, I will focus on what is worth predicting and for whom and how operational risks for the water system can be securitized while providing a platform for priced and negotiated allocation of the resources in the presence of imperfect forecasts. The possibility of a national or regional market for water contracts as part of the framework is explored, and its potential benefits and pitfalls identified.

Wednesday December 11, 2013

H31C. Managing and Modeling for Water Security I Posters [SWIRL_CU]
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“Assessment of Agricultural Water Management in Punjab, India using Bayesian Methods”
ID: 1806926

AUTHORS/INSTITUTIONS: T.A. Russo, N. Devineni, U. Lall, Columbia University, New York, New York, UNITED STATES;

R. Sidhu, Punjab Agricultural University, Ludhiana, Punjab, INDIA

BODY: The success of the Green Revolution in Punjab, India is threatened by the declining water table (approx. 1 m/yr). Punjab, a major agricultural supplier for the rest of India, supports irrigation with a canal system and groundwater, which is vastly over-exploited. Groundwater development in many districts is greater than 200% the annual recharge rate. The hydrologic data required to complete a mass-balance model are not available for this region, therefore we use Bayesian methods to estimate hydrologic properties and irrigation requirements. Using the known values of precipitation, total canal water delivery, crop yield, and water table elevation, we solve for each unknown parameter (often a coefficient) using a Markov chain Monte Carlo (MCMC) algorithm. Results provide regional estimates of irrigation requirements and groundwater recharge rates under observed climate conditions (1972 to 2002). Model results are used to estimate future water availability and demand to help inform agriculture management decisions under projected climate conditions. We find that changing cropping patterns for the region can maintain food production while balancing groundwater pumping with natural recharge. This computational method can be applied in data-scarce regions across the world, where agricultural water management is required to resolve competition between food security and changing resource availability.

H33M. Water Resources and Water Quality under Changing Climate and Land Use V
1:40 PM – 3:40 PM; 3016 (Moscone West)

“The effects of the African Green Revolution on nitrogen losses from two contrasting soil types in sub-Saharan Africa”
ID: 1803678

AUTHORS/INSTITUTIONS: K.L. Tully, T. Russo, J.E. Hickman, C. Palm, Earth Institute, Columbia University, New York, New York, UNITED STATES

BODY: Nearly 80% of countries in sub-Saharan Africa (SSA) face problems of nitrogen (N) scarcity, which together with poverty causes food insecurity and malnutrition. The Alliance for a Green Revolution in Africa has set a goal of increasing fertilizer use in the region six-fold by 2015. While there is substantial evidence that greater N fertilizer use will improve crop yields, it could lead to increased N leaching and elevated nitrate (NO3-) concentrations in surface water and groundwater reservoirs. However, it is unclear what the magnitude of impacts will be in SSA given historically low nutrient additions (of less than 5 kg N/ha/yr), highly degraded soils (due to years of nutrient and soil organic matter depletion), and a wide range of soil types on which increased fertilizer use is occurring. Current estimates of N dynamics and balances in SSA agriculture now rely on data from other regions with different soil types, soil fertility, and land management practices. To understand the influence of increased fertilizer use on water quality requires data from representative areas in SSA. Experimental maize plots were established in a randomized complete block design in both western Kenya (clayey soil) and mid-western Tanzania (sandy soil). Plots were amended with 0, 50, 75, and 200 kg N/ha/yr as mineral fertilizer. Tension lysimeters were installed at three depths in each treatment, and water was collected throughout the maize growing season. Soil water solutions were analyzed for NO3–N. Flow through the soil column at each soil depth, was modeled using VS2DT, a variably saturated flow and solute transport model, and water flux values were multiplied by measured NO3–N concentrations to estimate seasonal N leaching flux. Soil texture was a major driver of N losses, altering both the pathways and magnitude of losses. Clayey soils in western Kenya show an enormous potential for loss of NO3–N immediately following the onset of rains as they trigger high rates of N mineralization and nitrification in the topsoil (known as the “birch effect”). We did not observe this pulse in the sandy soils of central Tanzania. However, NO3- N concentrations in leachate were three times lower at 200 cm in clayey soils compared to sandy soils as a result of higher anion exchange capacity in clays. We show that while clayey soils lose NO3–N in a large pulse at the onset of rains, sandy soils lose large quantities of NO3–N over the course of the maize growing season. Results from this study can help inform recommended N application rates in similar soils (tropical Ultisols and Oxisols), to optimize yields while minimizing N leaching losses.

A31A. Fast Physics in Climate Models and Cloud-Resolving Models: Parameterization, Evaluation, and Observation I Posters (cosponsored by AMS)
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“Toward a unified stochastic representation of dry, shallow and deep convection”
ID: 1791678

AUTHORS/INSTITUTIONS: P. Gentine, F. D’Andrea, B.R. Lintner, A.K. Betts, Columbia University, New York, New York, UNITED STATES

BODY: A new bulk model for the convective boundary layer, the probabilistic bulk convection model (PBCM), is presented. PBCM implements a new closure based on the mass flux of f updrafts overshooting the inversion and originating from the surface. This mass flux is related to the subgrid scale variability of the surface state (potential temperature theta and specific humidity q) of a pdf of updraft plumes. Evaluating the model against observed clear-sky reference theta and q-profiles from the test cases of Sullivan (1998) and over the ARM SGP site shows that PBCM performs very well.

PBCM is extended to shallow convection regime. The cloud base mass flux is obtained explicitly by integrating the surface distribution over the uppermost buoyant updrafts. Significantly, the mass flux reduces the entrainment rate both directly, by reducing the number of updrafts contributing to the growth of the mixed layer, and indirectly, by inducing compensating subsidence on top of the mixed layer. Cloud cover is obtained naturally by integration over the condensating updrafts. Comparisons of PBCM cloud base, cloud top heights, cloud cover and cloud base mass flux against large-eddy simulations from the BOMEX and from the ARM SGP site demonstrate favorable agreement. At steady-state a tight equilibrium is naturally ensured between the cloud base mass flux, radiation and mixed-layer state.

The model is \ extended to deep convection to unveil the triggering mechanism of deep convection over land as well as the surface-moist convection feedbacks. Comparison of deep convection triggering over the African Monsoon Multidisciplinary Analysis (AMMA) campaign and ARM datasets demonstrate that PBCM is able to naturally characterize the diurnal cycle of deep convection triggering over continents. Extensions of the model toward a unified parameterization of convection will be discussed.

B33J. Postmortem: 2012 Drought—Terrestrial Ecosystems Posters
1:40 PM – 6:00 PM; Hall A-C (Moscone South)

“Survival strategies in semi-arid climate for isohydric and anisohydric species”
ID: 1818397

AUTHORS/INSTITUTIONS: M.F. Guerin, P. Gentine, Columbia University, New York, New York, UNITED STATES;

M. Uriarte, Ecology, Evolution & Env Biology, Columbia University, New York, New York, UNITED STATES

BODY: The understanding of survival strategies in dry land remains a challenging problem aiming at the interrelationship between local hydrology, plant physiology and climate. Carbon starvation and hydraulic failure are thought to be the two main factors leading to drought-induced mortality beside biotic perturbation. In order to better comprehend mortality the understanding of abiotic mechanisms triggering mortality is being studied in a tractable model for soil-plant-atmosphere continuum emphasizing the role of soil hydraulic properties, photosynthesis, embolism, leaf-gas exchange and climate. In particular the role of the frequency vs. the intensity of droughts is highlighted within such model. The analysis of the model included a differentiation between isohydric and anisohydric tree regulation and is supported by an extensive dataset of Pinion and Juniper growing in a semi-arid ecosystem. An objective of reduced number of parameters was approached with allometric equations to characterize tree’s main traits and their hydraulic controls. Leaf area, sapwood area and tree’s height are used to derive capacitance, conductance and photosynthetic abilities of the plant. A parameter sensitivity is performed highlighting the role of root:shoot ratio, rooting depth, photosynthetic capacity, quantum efficiency, and most importantly water use efficiency.

Analytic development emphasizes two regimes of transpiration/photosynthesis denoted as stage-I (no embolism) and stage-II (embolism dominated) in analogy with stage I-stage II treminology for evaporation (Phillip,1957). Anisohydric species tend to remain in stage-I during which they still can assimilate carbon at full potential thus avoiding carbon starvation. Isohydric species tend to remain longer in stage-II. The effects of drought intensity/frequency on those 2 stages are described.

While the field tests in Salvucci and Gentine (2013) supported the minimum variance hypothesis, the analysis did not reveal the mechanisms responsible for the behavior. Instead the paper suggested, heuristically, that the results were due to an equilibration of the relative humidity between the land surface and the surface layer of the boundary layer. Here we apply this method using surface meteorological fields simulated by a global climate model (GCM), and compare the predicted ET to that simulated by the climate model. Similar to the field tests, the GCM simulated ET is in agreement with that predicted by minimizing the profile relative humidity variance. A reasonable interpretation of these results is that the feedbacks responsible for the minimization of the profile relative humidity variance in nature are represented in the climate model. The climate model components, in particular the land surface model and boundary layer representation, can thus be analyzed in controlled numerical experiments to discern the specific processes leading to the observed behavior. Results of this analysis will be presented.

A31B. Dynamics and Predictability of Midlatitude Storms in a Changing Climate I Posters (cosponsored by AMS)
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“Equator-to-Pole and Ocean-Land Temperature Gradients in CMIP5 models and their Relation to Midlatitude Atmospheric Circulation”
ID: 1819775

AUTHORS/INSTITUTIONS: C. Karamperidou, Dept of Meteorology, University of Hawaii, Honolulu, Hawaii, UNITED STATES;
J.F. Booth, NASA GISS, New York, New York, UNITED STATES;
C. Karamperidou, U. Lall, Dept of Earth & Environmental Engineering, Columbia University, New York, New York, UNITED STATES

BODY: Zonal and meridional surface temperature gradients are associated with mean zonal circulation and storm-track variability. We investigate the projected changes in the equator-to-pole gradient and the ocean-land contrast in the suite of models participating in the Coupled Model Intercomparison Project 5 (CMIP5). The projected decrease in the gradients is in accordance with the poleward amplification of temperature increases and the faster warming of land compared to the ocean. However, we note differences in the projections between the Northern and the Southern Hemisphere, and among ocean basins in the Northern Hemisphere, that are indicative of competing mechanisms that underlie the projected shifts in the mean zonal circulation and storm tracks. As an additional indicator of the position and strength of midlatitude mean and transient atmospheric circulation, as well as the expansion of the tropics, we study the projected changes in the latitude of maximum meridional surface temperature gradient.

ED31C. Transformative Innovations in Earth, Oceans, and Atmospheric Science Education for Undergraduates Supported by the NSF-DUE Funding Programs, and Future Directions II Posters
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

“Adaptable Web Modules to Stimulate Active Learning in Engineering Hydrology using Data and Model Simulations of Three Regional Hydrologic Systems”
ID: 1792851

AUTHORS/INSTITUTIONS: E.H. Habib, M. Bodin, S. Chimmula, Department of Civil Engineering, University of Louisiana at Lafayette, Lafayette, Louisiana, UNITED STATES;
D.G. Tarboton, Utah Water Research Laboratory, Utah State University, Logan, Utah, UNITED STATES;
U. Lall, B. Rahill-Marier, Department of Earth and Environmental Engineering, Columbia University, New York City, New York, UNITED STATES;
E.A. Meselhe, The Water Institute of the Gulf, Baton Rouge, Louisiana, UNITED STATES;
A. Ali, South Florida Water Management District, West Palm Beach, Florida, UNITED STATES;
D. Williams, Y. Ma, Center for Innovative Learning and Assessment Technologies, University of Louisiana at Lafayette, Lafayette, Florida, UNITED STATES

BODY: The hydrologic community has long recognized the need for broad reform in hydrologic education. A paradigm shift is critically sought in undergraduate hydrology and water resource education by adopting context-rich, student-centered, and active learning strategies. Hydrologists currently deal with intricate issues rooted in complex natural ecosystems containing a multitude of interconnected processes. Advances in the multi-disciplinary field include observational settings such as Critical Zone and Water, Sustainability and Climate Observatories, Hydrologic Information Systems, instrumentation and modeling methods. These research advances theory and practices call for similar efforts and improvements in hydrologic education. The typical, text-book based approach in hydrologic education has focused on specific applications and/or unit processes associated with the hydrologic cycle with idealizations, rather than the contextual relations in the physical processes and the spatial and temporal dynamics connecting climate and ecosystems. An appreciation of the natural variability of these processes will lead to graduates with the ability to develop independent learning skills and understanding. This appreciation cannot be gained in curricula where field components such as observational and experimental data are deficient. These types of data are also critical when using simulation models to create environments that support this type of learning. Additional sources of observations in conjunction with models and field data are key to students understanding of the challenges associated with using models to represent such complex systems. Recent advances in scientific visualization and web-based technologies provide new opportunities for the development of active learning techniques utilizing ongoing research. The overall goal of the current study is to develop visual, case-based, data and simulation driven learning experiences to instructors and students through a web server-based system. Open source web technologies and community-based tools are used to facilitate wide dissemination and adaptation by diverse, independent institutions. The new hydrologic learning modules are based on recent developments in hydrologic modeling, data, and resources. The modules are embedded in three regional-scale ecosystems, Coastal Louisiana, Florida Everglades, and Utah Great Salt Lake Basin. These sites provide a wealth of hydrologic concepts and scenarios that can be used in most water resource and hydrology curricula. The study develops several learning modules based on the three hydro-systems covering subjects such as: water-budget analysis, effects of human and natural changes, climate-hydrology teleconnections, and water-resource management scenarios. The new developments include an instructional interface to give critical guidance and support to the learner and an instructor’s guide containing adaptation and implementation procedures to assist instructors in adopting and integrating the material into courses and provide a consistent experience. The design of the new hydrologic education developments will be transferable to independent institutions and adaptable both instructionally and technically through a server system capable of supporting additional developments by the educational community.

H33G. Regional Hydroclimatic Observations, Patterns, Changes, and Implications for Modeling II Posters
1:40 PM – 6:00 PM; Hall A-C (Moscone South)

Modeling Winter Rainfall in Northwest India using a Hidden Markov Model: Understanding Occurrence of Different States and their Dynamical Connections”
ID: 1819636

AUTHORS/ INSTITUTIONS :Indrani Pal,Department of Civil Engineering, University of Colorado, Denver, CO, United States, Andrew William Robertson,International Research Institute for Climate and Society, Columbia University, New York, NY, United States., Upmanu Lall, Department of Earth and Environmental Engineering, Columbia University, New York, NY, United States, Mark A Cane, Lamont-Doherty Earth Observatory, Columbia University, New York, NY, United States

BODY: A multiscale-modeling framework for daily rainfall is considered and diagnostic results are presented for an application to the winter season in Northwest India. The daily rainfall process is considered to follow a Hidden Markov Model (HMM), with the hidden states assumed to be an unknown random function of slowly varying climatic modulation of the winter jet stream and moisture transport dynamics. The data used are from 14 stations over the Satluj River basin in northwest India in winter (Dec-Jan-Feb-Mar). The period considered is 1977/78-2005/06. The HMM identifies four discrete weather states, which are used to describe daily rainfall variability over the study region. The first hidden state has low rainfall occurrence and intensity, the second has modest occurrence and low intensity, the third has high occurrence but low to modest intensity and the fourth has high frequency and intensity of daily rainfall. Each state was found to be associated with a distinct atmospheric circulation pattern, with States 3 and 4 characterized by a zonally oriented wave train extending across Eurasia between 20N–40N, identified with ‘Western Disturbances’. State 1, by contrast, is characterized by a lack of synoptic wave activity. The occurrence of State 4 is strongly conditioned by the El Nino and Indian Ocean Dipole (IOD) phenomena in winter, which is demonstrated using large-scale correlation maps based on mean sea level pressure (MSLP) and sea surface temperature (SST). This suggests that there is a tendency of higher frequency of the wet days and intense Western Disturbances in winter during El Nino and positive IOD years. These findings, derived from daily rainfall station records, help clarify the sequence of Northern Hemisphere mid-latitude storms bringing winter rainfall over Northwest India, and their association with potentially predictable low frequency modes on seasonal time scales and longer.

Thursday December 12, 2013

H41J. Statistical Modeling of Extreme Precipitation I Posters
8:00 AM – 12:20 PM; Hall A-C (Moscone South)

A Nonparametric Simulator for Multivariate Random Variables with Differing Marginal Densities and Non-linear Dependence with Hydroclimatic Applications”

ID: 1811064
AUTHORS/ INSTITUTIONS: D. Farnham, U. Lall, B. Rahill-Marier, Earth and Environmental Engineering, Columbia University, New York, New York, UNITED STATES;

U. Lall, N. Devineni, Columbia Water Center, New York, New York, UNITED STATES

BODY: Hydrologic models often require as inputs stochastic simulations of meteorological variables that are mutually consistent and spatially coherent, i.e., have marginal and joint probability densities that correspond to those estimated from physically realized states, and have the appropriate spatial structure. These inputs may come from historical meteorological data, or from relatively small ensembles of integrations of numerical climate and weather models. Often, empirical modeling or simulation of multiple hydroclimatic variables, or simulations of hydrologic variables at multiple sites that respect the spatial co-variability may also be desired. A nonparametric simulation strategy is presented that is capable of 1) addressing marginal probability density functions that are different for each variable of interest and 2) reproducing the joint probability distribution across a potentially large set of variables or spatial instances. The application of rainfall simulations developed from historic rain gauge and radar data is explored. Such simulations are useful for urban hydrological modelers seeking more spatially resolved precipitation forcings.

 H43N. Statistical Modeling of Extreme Precipitation II
1:40 PM – 3:40 PM; 3011 (Moscone West)

“Multivariate Bayesian Models of Extreme Rainfall”
ID: 1809248

AUTHORS/INSTITUTIONS: B. Rahill-Marier, U. Lall, D. Farnham, Earth & Environmental Engineering, Columbia University, New York, New York, UNITED STATES;
B. Rahill-Marier, N. Devineni, U. Lall, D. Farnham, Columbia Water Center, New York, New York, UNITED STATES

BODY: Accounting for spatial heterogeneity in extreme rainfall has important ramifications in hydrological design and climate models alike. Traditional methods, including areal reduction factors and kriging, are sensitive to catchment shape assumptions and return periods, and do not explicitly model spatial dependence between between data points. More recent spatially dense rainfall simulators depend on newer data sources such as radar and may struggle to reproduce extremes because of physical assumptions in the model and short historical records. Rain gauges offer the longest historical record, key when considering rainfall extremes and changes over time, and particularly relevant in today’s environment of designing for climate change. In this paper we propose a probabilistic approach of accounting for spatial dependence using the lengthy but spatially disparate hourly rainfall network in the greater New York City area. We build a hierarchical Bayesian model allowing extremes at one station to co-vary with concurrent rainfall fields occurring at other stations. Subsequently we pool across the extreme rainfall fields of all stations, and demonstrate that the expected catchment-wide events are significantly lower when considering spatial fields instead of maxima-only fields. We additionally demonstrate the importance of using concurrent spatial fields, rather than annual maxima, in producing covariance matrices that describe true storm dynamics. This approach is also unique in that it considers short duration storms – from one hour to twenty-four hours – rather than the daily values typically derived from rainfall gauges. The same methodology can be extended to include the radar fields available in the past decade. The hierarchical multilevel approach lends itself easily to integration of long-record parameters and short-record parameters at a station or regional level. In addition climate covariates can be introduced to support the relationship of spatial covariance with meteorological phenomena as well as long-term climate trends.

A43B. Cloud, Convection, Radiation, Water and Energy Cycles I Posters
1:40 PM – 6:00 PM; Hall A-C (Moscone South)

“Continental radiative-convective equilibrium experiments in a single column model (LMDZ5B GCM)”
ID: 1792877

AUTHORS/INSTITUTIONS: N. Rochetin, P. Gentine, A.H. Sobel, Columbia University in the City of New York, New York, New York, UNITED STATES;
B.R. Lintner, Rutgers University, Rutgers, New Jersey, UNITED STATES;
K.L. Findell, GFDL, Princeton University, Princeton, New Jersey, UNITED STATES

BODY: The radiative-convective instability results both from (i) the average net cooling experienced by the Earth’s atmosphere (~ 110 W/m?) and (ii) from the equivalent warming of the Earth’s surface. Ultimately, this drives the Earth atmosphere to a radiative-convective equilibrium (RCE) state, in a sense that, at the global scale, surface fluxes and radiative cooling compensate each other. Since the convection time scale (i.e. some hours) is much shorter than the radiation one (i.e. about 40 days), the resulting global temperature lapse rate is generally closer to the moist adiabat than to the dry adiabat. This is especially true over the tropics, where moist convection is in near-equilibrium. The RCE is then often used as a common approximation of the tropical mean state. It has been extensively used over oceans in SCMs (Single Column Models), as well as in CRMs (Cloud Resolving Models), to investigate the tropical moist convection sensitivity (i) to boundary conditions (e.g. SST, surface wind, drag coefficient, etc…) and (ii) to atmospheric conditions (e.g. radiative cooling, wind shear, tropospheric humidity, etc…).

Nevertheless, to our knowledge the present study is the first one investigating the RCE over a continental surface. Indeed, in the present study, the single column version of the LMDZ GCM (LMDZ5B, from the Laboratoire de Meteorologie Dynamique) is ran to RCE, with a coupled land surface both in terms of temperature and moisture. This continental RCE demonstrates very different sensitivity compared to its oceanic counterpart in particular because of the large- amplitude heat flux diurnal cycle, which is shown to strongly impact the equilibrium state.

Sensitivity studies (i) to solar forcing (latitude) (ii) to total water content, and (iii) to the initial conditions are performed to study the different equilibrium states, with a particular focus on the role of clouds. We also performed a bifurcation diagram. Low-level clouds and fog are shown to be key for the equilibrium since they strongly modify the diurnal course of heat at the land surface. Total water content is shown to mostly impact the soil water budget but, interestingly, the surface energy budget and atmospheric moisture is not too much sensitive to the initial total water content; that is, for a given latitude, the atmospheric precipitable water tends toward a single attractor.

H43O. Water Resources Management and Policy in a Changing World II: Tackling Water Problems Around the Globe
1:40 PM – 3:40 PM; 3016 (Moscone West)

“Averting a Disaster with Groundwater Depletion in India: The General Case of Water Management Principles and Development”
ID: 1808048

AUTHORS/INSTITUTIONS: U. Lall, Earth & Env Eng, Columbia Univ, New York, New York, UNITED STATES

BODY: Many countries, including the USA, China, and India are experiencing chronic groundwater depletion. In part this unsustainable water use results from climatic factors that reduce surface water availability and also the recharge to the aquifer system. However, a more critical factor is uncontrolled use for agriculture and energy and mineral processing. Interestingly in places such as India endowments have been politically created that lead to ever increasing use, through the provision of free energy for pumping. Reversing the situation is considered politically challenging, and the concept of metering and payment for what is essentially economic use of water is also considered difficult to apply.

In this talk I use the Indian situation as a general example and discuss the role central planning strategies for demand and resource management can play recognizing the private action by millions of users as an inevitable tool that needs to be leveraged without necessarily the high transaction costs that come with monitoring and fee collection for monitored use. Specifically, targeting and stimulating potential cropping strategies and on farm water and energy management emerge as a choice in a difficult management environment. In a broader development context, I argue that the role of private sector aggregators in developing farm to market procurement strategies can play a role in both improving rural economies and providing a trajectory for more efficient water use through technology and crop choice.

GC43C. General Circulation Model Downscaling for Impact, Vulnerability and Adaption Assessments: Methodologies and Applications II Posters [SWIRL_CM]
1:40 PM – 6:00 PM; Hall A-C (Moscone South)

“Projecting Changes in S. Florida Rainfall for the 21st century: Scenarios, Downscaling and Analysis”
ID: 1807938

AUTHORS/INSTITUTIONS: F. Cioffi, A. Monti, University of Rome La Sapienza, Rome, ITALY;

F. Cioffi, U. Lall, Earth and Environmental Engineering, Columbia University, New York, New York, UNITED STATES

BODY: A Non-Homogeneous hidden Markov Models (NHMM) is developed using a 65-years record (1948-2012) of daily rainfall amount at nineteen stations in South Florida and re-analysis atmospheric fields of Temperature (T) at 1000 hPa, Geo Potential Height (GPH) at 1000 hPa, Meridional Winds (MW) and Zonal Winds (ZW) at 850 hPa, and Zonal Winds on the specific latitude of 27N (ZW27N) from 10 to 1000 hPa.

The NHMM fitted is then used for predicting future rainfall patterns under global warming scenario (RCP8.5), using predictors from the CMCC-CMS simulations from 1950-2100. The model directly includes a consideration of seasonality through changes in the driving variables thus addressing the question of how future changes in seasonality of precipitation can also be modeled.

The results of the simulations obtained by using the downscaling model NHMM, with predictors derived from the simulations of CMCC-CMS CGM, in the worst conditions of global warming as simulated by RCP8.5 scenario, seems to indicate that, as a consequence of increase of CO2 concentration and temperature, South Florida should be subjected to more frequent dry conditions for the most part of the year, due mainly to a reduction of number of wet days and, at the same time, the territory should be also affected by extreme rainfall events that are more intense than the present ones. What appears from results is an increases of rainfall variability.

This scenario seems coherent with the trends of rainfall patterns observed in the XX century. An investigation on the causes of such hydrologic changes, and specifically on the role of North Atlantic Subtropical High is pursued.

Friday December 13, 2013

H51Q. Hydroclimatic Extremes: Estimation and Forecasting III (Virtual Option)
8:00 AM – 10:00 AM; 3002 (Moscone West)

“Improving Local and Regional Flood Quantile Estimates Using a Hierarchical Bayesian GEV Model”
ID: 1810309

AUTHORS/INSTITUTIONS: C.H. Ribeiro Lima, Civil and Environmental Engineering, University of Brasilia, Brasilia, BRAZIL;

C.H. Ribeiro Lima, U. Lall, N. Devineni, T. Troy, Water Center, Columbia University, New York, New York, UNITED STATES

BODY: Flood risk management usually relies on local and regional flood frequency analysis, which tends to suffer from lack of data and parameter uncertainties. Here we estimate local and regional Generalized Extreme Value (GEV) distribution parameters in a hierarchical Bayesian framework, which helps reduce uncertainties by pooling more information in the estimation process and provides a simple topology to propagate model and parameter uncertainties to flood quantile estimates. As prior information for the Bayesian model, it is assumed for each site that the GEV location and scale parameters come from independent log-normal distributions, whose mean parameter follows the well known log-log scaling law with the drainage area. The shape parameter for each site is shrunk towards a common mean. Non-informative prior distributions are assumed for the hyperparameters and the MCMC method is used to sample from the posterior distributions. The model is tested using annual maximum series from 20 streamflow gauges located in an 83.000 km2 basin in southeastern Brazil. The results show a significant improvement of flood quantile estimates over the traditional GEV model, particularly for sites with few data. For return periods within the range of the data (around 50 years), the Bayesian credible intervals for the flood quantiles are narrower than the classical confidence limits based on the delta method. As the return period increases beyond the range of the data, the confidence limits from the delta method become unreliable and the Bayesian credible intervals provide a way to estimate satisfactory confidence bands for the flood quantiles considering the parameter uncertainties. In order to evaluate the applicability of the proposed hierarchical Bayesian model for flood frequency regional analysis, we estimate flood quantiles for three randomly chosen out-of-sample sites and compare with classical estimates using the index flood method. The posterior distributions of the scaling law coefficients are used to define the predictive distributions of the GEV location and scale parameters for the out-of-sample sites given only their drainage areas and the posterior distribution of the average shape parameter is taken as the regional predictive distribution for this parameter. While the index flood method does not provide a straightforward way to consider the uncertainties in the index flood and in the regional parameters, the results obtained here show that the proposed Bayesian method is able to produce adequate credible intervals for flood quantiles that are in accordance with empirical estimates.

H53D. Global Floods: Satellite Observation, Modeling and Socioeconomic Response II Posters [SWIRL_CM]
1:40 PM – 6:00 PM; Hall A-C (Moscone South)

“Climate-informed flood risk estimation”
ID: 1815348

AUTHORS/INSTITUTIONS: T. Troy, N. Devineni, U. Lall, Columbia Water Center, The Earth Institute at Columbia University, New York, New York, UNITED STATES;
T. Troy, Civil and Environmental Engineering, Lehigh University, Bethlehem, Pennsylvania, UNITED STATES;
C. Lima, Universidade de Brasilia, Brasilia, BRAZIL;
U. Lall, Earth and Environmental Engineering, Columbia University, New York, New York, UNITED STATES

BODY: Currently, flood risk assessments are typically tied to a peak flow event that has an associated return period and inundation extent. This method is convenient: based on a historical record of annual maximum flows, a return period can be calculated with some assumptions about the probability distribution and stationarity. It is also problematic in its stationarity assumption, reliance on relatively short records, and treating flooding as a random event disconnected from large-scale climate processes. Recognizing these limitations, we have developed a new approach to flood risk assessment that connects climate variability, precipitation dynamics, and flood modeling to estimate the likelihood of flooding.

To provide more robust, long time series of precipitation, we used stochastic weather generator models to simulate the rainfall fields. The method uses a k-nearest neighbor resampling algorithm in conjunction with a non-parametric empirical copulas based simulation strategy to reproduce the temporal and spatial dynamics, respectively. Climate patterns inform the likelihood of heavy rainfall in the model. For example, ENSO affects the likelihood of wet or dry years in Australia, and this is incorporated in the model. The stochastic simulations are then used to drive a cascade of models to predict flood inundation. Runoff is generated by the Variable Infiltration Capacity (VIC) model, fed into a full kinematic wave routing model at high resolution, and the kinematic wave is used as a boundary condition to predict flood inundation using a coupled storage cell model. Combining the strengths of a stochastic model for rainfall and a physical model for flood prediction allows us to overcome the limitations of traditional flood risk assessment and provide robust estimates of flood risk.

H52G. Water Resources Management and Policy in a Changing World V: Dealing with Flow and Climate Variability
10:20 AM – 12:20 PM; 3016 (Moscone West)

“Multi-time scale Climate Informed Stochastic Hybrid Simulation-Optimization Model (McISH model) for Multi-Purpose Reservoir System”
ID: 1787528

AUTHORS/INSTITUTIONS: M. Lu, U. Lall, Earth & Environmental Engineering, Columbia University in the City of New York, New York, New York, UNITED STATES;
U. Lall, International Research Institute for Climate & Society, Palisades, New York, UNITED STATES

BODY: In order to mitigate the impacts of climate change, proactive management strategies to operate reservoirs and dams are needed. A multi-time scale climate informed stochastic model is developed to optimize the operations for a multi-purpose single reservoir by simulating decadal, interannual, seasonal and sub-seasonal variability. We apply the model to a setting motivated by the largest multi-purpose dam in N. India, the Bhakhra reservoir on the Sutlej River, a tributary of the Indus. This leads to a focus on timing and amplitude of the flows for the monsoon and snowmelt periods. The flow simulations are constrained by multiple sources of historical data and GCM future projections, that are being developed through a NSF funded project titled “Decadal Prediction and Stochastic Simulation of Hydroclimate Over Monsoon Asia”.

The model presented is a multilevel, nonlinear programming model that aims to optimize the reservoir operating policy on a decadal horizon and the operation strategy on an updated annual basis. The model is hierarchical, in terms of having a structure that two optimization models designated for different time scales are nested as a matryoshka doll. The two optimization models have similar mathematical formulations with some modifications to meet the constraints within that time frame. The first level of the model is designated to provide optimization solution for policy makers to determine contracted annual releases to different uses with a prescribed reliability; the second level is a within-the-period (e.g., year) operation optimization scheme that allocates the contracted annual releases on a subperiod (e.g. monthly) basis, with additional benefit for extra release and penalty for failure. The model maximizes the net benefit of irrigation, hydropower generation and flood control in each of the periods.

The model design thus facilitates the consistent application of weather and climate forecasts to improve operations of reservoir systems. The decadal flow simulations are re-initialized every year with updated climate projections to improve the reliability of the operation rules for the next year, within which the seasonal operation strategies are nested. The multi-level structure can be repeated for monthly operation with weekly subperiods to take advantage of evolving weather forecasts and seasonal climate forecasts. As a result of the hierarchical structure, sub-seasonal even weather time scale updates and adjustment can be achieved. Given an ensemble of these scenarios, the McISH reservoir simulation-optimization model is able to derive the desired reservoir storage levels, including minimum and maximum, as a function of calendar date, and the associated release patterns. The multi-time scale approach allows adaptive management of water supplies acknowledging the changing risks, meeting both the objectives over the decade in expected value and controlling the near term and planning period risk through probabilistic reliability constraints. For the applications presented, the target season is the monsoon season from June to September. The model also includes a monthly flood volume forecast model, based on a Copula density fit to the monthly flow and the flood volume flow. This is used to guide dynamic allocation of the flood control volume given the forecasts.

Submit Comment

Your email address will not be published. Required fields are marked *