H13B-0912
Developing a Framework for Assessing Climate-Change Impacts on Common Loon Habitat Suitability in Northern Wisconsin
A major focus of the U.S. Geological Survey's Trout Lake Water, Energy and Biogeochemical Budgets (WEBB) project is the development of a watershed model to allow predictions of hydrologic response to future conditions including land-use and climate. The coupled ground-water/surface-water model GSFLOW was chosen for this purpose because it could easily incorporate an existing ground-water flow model and it provides for simulation of surface-water processes. The Trout Lake watershed in Northern Wisconsin is underlain by a highly conductive glacial outwash sand aquifer. In this area, stream flow is dominated by ground-water contributions; however, surface runoff does occur during intense rainfall periods and spring snowmelt, and locally in near-stream/lake areas where the unsaturated zone is thin. Model calibration was performed using the PEST automated parameter estimation suite of software and the time-series processing utility. PEST was also used to provide estimates of uncertainty associated with the model predictions. The calibrated model was used to simulate the hydrologic response of the study lakes to a variety of climate-change scenarios culled from the IPCC Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Results from the simulations indicate climate change could result in substantial changes to the hydrologic budgets of the selected study lakes. A framework of models is being constructed to assess the impact of climate change on loon nesting. Previous research indicates several physical attributes of lakes, including water chemistry and clarity, are associated with breeding territory selection by Wisconsin common loons. Results from the hydrologic simulations, along with particle tracking within the ground-water portion of the flow model, will be used to develop predictions of solutes into the selected lakes. Estimates of solute concentrations within the ground- water system will be based on relating flow-path residence time to concentrations based on geochemical modeling of the system. The DYRESM-CAEDYM lake model will then be used to predict changes in the key water-chemistry measures needed by the loon habitat model.
H13B-0913
Multiple Hydrological Attractors and the Probability of Threshold Crossing
A numerical distributed eco-hydrological model (coupled hillslope Boussinesq-vertically lumped vadose zone)
has been developed in which qualitatively different steady state (attractor) water table elevations (shallow
and deep) exist for the same parameter set and is used here to analyse catchment resilience to climatic
disturbances. Traditional resilience investigations use limit cycle continuation (LCC) to quantified the state
space location and number of attractor and thresholds (repellor). A deficiency of this method is the
requirement to use constant or stable and smooth inter-annual cycles of climate forcing. As infiltration and
recharge are dominated by climatic events, rather than monthly averages, the use of LCC is questionable.
To overcome this limitation, simulations were instead undertaken with 100 stochastic climate sets for a range
of saturated conductivity parameter (ksat) values to assess if and at what parameter values i) bimodal
water table elevations emerge thus indicating multiple attractors; and ii) too quantify the probability of a shift
from a shallow to a deep water table attractor, and vice versa. Both monthly and daily climate data were
investigated.
To assess bimodal behaviour, for each ksat value the mean depth to the water table over the final year
of simulation was calculated and a histogram derived from the 100 simulations. The emergence of multiple
modes was found to be dependent upon the spatially-averaged limiting infiltration rate parameter. For a high
limiting infiltration rate (20 mm/day) the system shifts from the deep to the shallow state at the first major
rainfall event thus resulting in only one clear mode. Reducing the rate by 50% and 75% resulted in
increasingly clearer bimodal water table elevations but over a lesser range of ksat than that from the
LCC. To simplify histogram results to a form more comparable to the LCC, the probability of an attractor shift
was quantified. The probability of a shift was found to differ significantly depending upon whether the initial
state was the shallow or deep water table attractor, with a shift from the deep to the shallow state being was
more probable than the reverse. This difference in resilience could not have been identified from LCC as the
state space distance from each attractor to the repellor was near equal. The significance of this work is in i)
the further development of techniques for quantifying resilience; ii) strengthening the theory of multiple
attractors of dryland catchments; and iii) improving the characterisation of which catchments are more likely
to have multiple attractors.
H13B-0914
Land Surface Processes and Vegetation Interaction in a Changing Climate
The coupled interaction between hydrological processes and vegetation seasonality is regarded as fundamental for enhancing predictability of surface energy fluxes and the soil moisture dynamics. Investigation of this interaction becomes more important in a changing climate, as non-stationarity in the climate system may induce changes that can amplify or reduce the feedbacks between hydrological and vegetation processes. This study designs a "proof-of-concept" framework allowing one to investigate the propagation of climate disturbances, as inferred, for example, from global circulation models, through the eco-hydrologic system. Two different scenarios are discussed. The first one reproduces current climate and the second one generates future climates based on parameterizations developed in a weather generator. The weather generator reproduces climate statistics over a wide range of temporal scales, from the high- frequency extremes, to the low-frequency interannual variability. The generator has been designed to simulate input variables for an ecohydrological model at the hourly scale. The hydro-climatological variables obtained for the two scenarios serve as input to a hydrological model coupled with a model of vegetation dynamics. The latter model parsimoniously parameterizes essential plant life-cycle processes, including phenology. A comparison between the two cases in terms of vegetation response and hydrological states are examined, discussing possible implication of the feedbacks between ecosystem and hydrological cycle in a non-stationary climate
H13B-0915
Carbon Dynamics of Montane Native Hawaiian Rainforests Under Climate Change: Empirical and Modeling Studies
Recent studies (e.g., Giambelluca et al., 2008) have shown that Hawaiian montane air temperature has increased over 0.8 degree C in the past three decades, with the most pronounced increases observed in minimum nighttime temperatures and a decrease in the amplitude of the diurnal temperature cycle. Yet the effects of such changes on Hawaiian Ecosystems remain unknown. In this study, we investigate the impact of climate change on native Metrosideros polymorpha rainforests using both empirical and modeling approaches. We have collected eddy covariance measurements over a period of 3.5 years in a native M. polymorpha forest at a 1219-m elevation, windward site on the Island of Hawai'i. We found that each degree C increase in air temperature causes a 15% increase in respiration, which suggests that these montane rainforests may be negatively affected by future warming (Asner et al., manuscript). To confirm our findings, we plan to extend our study to a longer period, using the combined Simple Biosphere/Carnegie-Ames-Stanford Approach (SiBCASA) terrestrial carbon cycle model (Schaefer et al., 2008). We expect to achieve a better understanding of the carbon dynamics and its interaction with hydrology of Montane Native Hawaiian rainforest under climate change through the modeling study. The model will be calibrated and validated with the eddy covariance measurements.
H13B-0916
Investigating Vegetation Dynamics When Scaled From Plant to Ecosystems
As we scale up biophysical processes from plant to ecosystems, several important biotic and abiotic interactions need to be captured in a model. The biotic heterogeneity of plant species composition and distribution, and changes there in due to fluctuation in resources such as light, water, nutrients etc. arising from growth, disturbance, recruitment and mortality play an important role in the scaling up of fluxes. Even under homogeneous abiotic environments, there are inter-plant interactions which can be both competitive and cooperative and this leads to horizontal fluxes. Ecosystem Demography (ED) model [Moorcroft, 2001] provides an attractive framework to capture these behavior using a stochastic-dynamic approach. In this work, we embed a multilayer vegetation and soil model within the ED model to understand the scaling up dynamics. This approach allows us to capture plant acclimation due to changes in ambient CO2 levels, hydraulic redistribution by roots, and inter and intra species competition for resources. This model is used to predict the effect of climate change on the growth and dynamics of plants at the ecosystem scale. In particular, we study the emergent properties in terms of plant acclimation strategies at the ecosystem scale. These results are investigated to study regime shifts and changes in ecosystem functionality and composition due to climate change. An improved carbon allocation model helps us investigate above and below ground plant biomass (carbon and nitrogen) at different scales and its effect on water, carbon and nitrogen cycles. These investigations provide valuable insights into the understanding of climate change on vegetation and the associated feedback across a range of scales. Moorcroft PR, Hurtt GC, Pacala SW "A method for scaling vegetation dynamics: The ecosystem demography model (ED)" Ecological Monographs. Vol. 71: 4, p 557-585, 2001
H13B-0917
Scaling Leaves to Fluxes in a Well-Watered Saline Soil Peatlands
This study investigates the role of soil salinity in modifying land surface fluxes in a California Bay-Delta peatlands pasture. It combines leaf gas exchange and soil salinity measurements in a soil-vegetation- atmosphere transfer model to move from the plant scale to the footprint of an eddy covariance tower. The close match between measured and modeled daytime carbon and water fluxes encourages numerical analyses and predictions for assessing the separate and joint effects of various drivers of and responses to climate change. The model examines the role of temperature, carbon dioxide concentration, soil water content, and soil salinity levels on the fundamental fluxes of carbon and water between the surface and the atmosphere. Incorporating both leaf gas exchange parameters and responses to soil salinity in the model are shown to be critical to accurate assessments of the diurnal cycles of the fluxes. Only by incorporating both sets of knowledge into the model are proper estimates of water use efficiency guaranteed. The model and data reveal, through rigorous sensitivity analysis, the resilience of the dominant landscape cover, Lepidium latifolium, to the drivers of and responses to climate change. This research is useful to regional water resources and land management planners, as the landscape is capable of generating very high evapotranspiration rates (up to 5 mm per day). Moreover, this land cover species is considered by the California Department of Agriculture a species of concern due to its ability to expand widely and create dense mono-specific stands that exclude other plants and impact the native ecological and agricultural plant communities. Understanding its response to climate change will be of critical importance in accurately assessing its control.
H13B-0918
Scaling rules for understanding coupled carbon and water cycling responses to hydrologic and vegetation changes in semiarid ecosystems
An overriding challenge to understand and predict the resilience of coupled ecosystem processes is to generate robust and mechanistic scaling rules. Semiarid ecosystems have a tight coupling between carbon and water cycling and scaling rules need to mechanistically translate these couplings from individual leaves to landscapes. We are developing such rules in the context of asking how does precipitation differentially affect phenological timing and metabolic activities in southern Arizona landscapes? We compared phenological variation through overlays of remotely sensed vegetation indices and modeled meteorology between 39 distinct 6.25 km2 landscapes distributed in southern Arizona for seven years. Regression analyses were used to identify relationships between landscape co-variability in phenology and precipitation. Both summer and winter peak vegetation responded positively to precipitation; however, the timing of phenological events, interactions with prior season precipitation, and spatio-temporal variability varied substantially between seasons. Surprisingly, summer maximum growth was negatively affected by winter season precipitation. We compared metabolic dynamics and their linkages to water cycling through a data- model assimilation analysis that combined of a network of environmental observatories with metabolic models of respiration and photosynthesis. Using a metabolic framework, we described precipitation induced pulses as a reduction in the activation energy for respiration after individual precipitation events. Over the 5 day post-event period across four contrasting sites the mean precipitation effect was attributed to 6.1 g CO2 m-2 of carbon release, which represented a 21% increase in respiration over the pre-event steady state trajectory of carbon loss. To describe carbon uptake rates we examined a simplified sub-hourly ecosystem physiologic theory based on fundamental principles of diffusion, mass balance, reaction kinetics, and biochemical regulation of photosynthesis. An initial data-model fusion product comparing a woodland and grassland site skillfully reproduced the majority of diurnal variation in NEE for all sampling periods. The woodland site had consistently higher assimilation rates, lower seasonal variability, and a larger diurnal assimilation hysteresis compared to the grassland site. The largest proportion of variation in assimilation and net exchanges was associated with physiological differences. When comparing the two community types, the woodland showed a greater sensitivity than the grassland to water supply, while the grassland showed a greater sensitivity to leaf area than the woodland. We are currently coupling these two sources of information, intensive metabolic and extensive phenological, to estimate spatial patterns and regional budgets in the coupling between carbon and water cycles for the southern Arizona. These spatially explicit characterizations of coupled carbon and water dynamics will help identify locations vulnerable to future meteorological perturbations.
H13B-0919
The Transformation of a Semiarid Ecosystem Due to Severe Drought and How It Has Influenced the Hydrologic Cycle Across Varying Scales
An extended, severe drought in the southwestern U.S. from 2000 to 2003 was accompanied by increased temperatures and bark beetle infestations, inducing the large-scale mortality of woody overstory (Pinus edulis). The consequential redistribution of water, radiation, and nutrient availability modified the ecosystem phenology, species composition, and forced the ecosystem to transition into a new state. We hypothesize that the hydrological processes in the ecosystem were also altered due to the mortality. Thus, our objective is to investigate changes in the soil-vegetation-atmosphere continuum across the plot, watershed, and ecoregion scales. The plot site is located near Los Alamos in Northern New Mexico (1.5 hectare), the watershed is the Rio Ojo Caliente Basin (1,050 km2), a subbasin of the Upper Rio Grande, and the ecoregion consists of Pinus edulis, or piñon, across the Four Corners Region of Arizona, Colorado, New Mexico, and Utah (245,000 km2). These sites are chosen because a significant portion of the woodland ecosystem (piñon-juniper) was affected during the mortality event. Examining a remotely-sensed vegetation index (1-km AVHRR NDVI from 1989 to 2007), there is an increasing trend in the NDVI from 1989 to 1998 (pre-drought period), a decreasing trend from 1999 to 2003 (drought period), and a dramatic increasing trend from 2004 to 2007 (post-drought period) in which the NDVI rebounds to nearly pre-drought magnitudes. This pattern exists across the three spatial scales and signifies a profound alteration in the ecosystem, for while the vegetation composition was altered to a great degree, the system rapidly recovered photosynthetically during the post-drought period. This may be attributable to the decrease in the less- responsive overstory (pinñon mortality) and increase in the more-responsive understory (grasses and shrubs exploiting newly available resources). In order to examine hydrological changes, temporal patterns in gauge-based precipitation (frozen and unfrozen) and air temperature, and spatial-temporal patterns in PRISM precipitation, air temperature, and a soil moisture index are compared to the NDVI. The aim of this research is to explore the consequences of a severe drought married with elevated temperatures on vegetation and water resources. As the intensity and frequency of droughts are expected to increase in the southwestern U.S. with rising temperatures (IPCC 2007), this research contributes to our knowledge of ecosystem and hydrologic response to the changing climate.
H13B-0920
Improving Riparian Land Surface Flux Estimation Through Root Zone Groundwater Interaction in Semiarid Environments
Accurate estimation of turbulent land surface fluxes in land surface models (LSM) is important for sustainability studies of riparian areas, e.g. how they react to changes in groundwater availability either through extended droughts or due to climate change. Because atmospheric evaporative demand is very high in semiarid environments, neglecting important sources of root water uptake can cause errors in the estimation of these fluxes. In this study we apply a simple parameterization of root-zone groundwater interaction to the Noah LSM and compare coupled (to groundwater) and uncoupled model runs to in situ observations of latent heat fluxes and soil moisture. Observations used in this study were collected at a riparian grassland site located near the San Pedro River, in Arizona, USA. Our results show that when the Noah model is run without the presence of a water table (uncoupled run) we get significant under prediction of the latent heat fluxes, and by introducing a root-zone groundwater interaction (coupled run) we are able to better capture the observed turbulent flux dynamics. However, the coupled model still does not predict observed soil moisture dynamics accurately and in general overestimates soil moisture in the upper soil layers. Our findings suggest that a different root water uptake mechanism than currently implemented in Noah may be required to simultaneously reproduce both the latent heat and root zone water balance dynamics.
H13B-0921
Soil Physical Properties Modulate the Effects of Climate Variability on Aboveground Productivity in a Semiarid Rangeland
Accurate characterization of soil-water variability and its relation to ecosystem primary productivity are critical to understanding ecosystem response to climate change. We hypothesized that soil development modulates climate variability and controls primary productivity through control of soil-water availability, residence time, and temporal variation. This hypothesis was tested in a semiarid rangeland in the US Southwest by coupling a high temporal resolution meteorological and soil-water dataset from Holocene and Pleistocene aged soils with remotely sensed metrics of aboveground productivity. Holocene and Pleistocene soils varied significantly in soil physical properties such as clay and organic matter content and bulk density. A multi-year time series of Normalized Difference Vegetation Index (NDVI) data indicated distinct patterns in aboveground production by landscape age; both the start of growing season and peak NDVI values of Pleistocene soils lagged behind Holocene soils. In addition, Pleistocene soil aboveground biomass productivity was roughly 75% that of Holocene soils. The link between soil-water and NDVI was carried out using simulations of root water uptake in HYDRUS-1D. Soil hydraulic properties were generated with Rosetta and HYDRUS simulations provisionally matched to observed soil-water contents. Pleistocene soils exhibited an order of magnitude lower saturated hydraulic conductivity as well as less potential soil-water availability, relative to Holocene soils. Plant transpiration estimated from HYDRUS demonstrated a high degree of correspondence to aboveground productivity patterns observed with integrated NDVI values for both soils. Correlations (R) between NDVI and modeled transpiration data were 0.79 and 0.89 for the Pleistocene and Holocene data, respectively, indicating a very strong correspondence between NDVI and soil-water status. These data indicate that variation in aboveground productivity with landscape age may be quantitatively linked to soil hydraulic properties and soil-water dynamics.
H13B-0922
Interannual and Intraseasonal Interactions Between Greening Process and Soil Moisture in the North American Monsoon Region in Northwestern Mexico
In the North American Monsoon System (NAMS) Tier 1 core area, summer precipitation accounts for more than fifty percent of the annual total. Early in this season precipitation events produce rapid greening of the vegetation cover. While this greening has been shown to results in major changes in land surface water and energy cycles locally, the nature of the resulting interactions with regional scale fluxes, and its intraseasonal and interannual variability, are less clear. We focus here on the greening process and soil moisture relationships at the regional scale, as key components of the hydrological cycle associated with the greening. Specifically, using MODIS LAI from 2000 to 2007, we identify spatiotemporal patterns of variability of the greening process. We also evaluate assimilation of MODIS LAI into the Variable Infiltration Capacity (VIC) land surface model in order to estimate the implications of interannual variations in vegetation greenness on surface hydrological conditions. In these experiments, we force VIC with daily observed precipitation, minimum and maximum temperatures, and wind speed at 1/16 degree spatial resolution over the NAME tier 1 core area in northwest Mexico. We use VIC soil moisture estimates both with a fixed seasonal cycle of leaf area index, and with MODIS LAI, to evaluate interannual differences in the hydrological response to changes in vegetation during the pre-monsoon and monsoon months of wet and dry years and during the specific greening events between 2000 and 2007. These experiments allow us to evaluate the role of soil moisture persistence on the intraseasonal and interannual variability of the greening process and to investigate factors that may affect the strength of the greening-soil moisture relationship.
H13B-0923
Effects of Precipitation Patterns on Productivity of a Semi-arid Forest
Global climate changes are predicted to be associated with both reduced total precipitation and increased storm intensities over many regions, including the entire Mediterranean. The combined effect is difficult to predict and can greatly benefit from studies in dry forest ecosystems exposed to large inter-annual climatic variations. We report on a four year study (2003-2007) in the Yatir semi-arid pine forest in Southern Israel (40 years old P. halepensis; LAI=1.5; mean precipitation 280 mm/yr) that showed resistance to seasonal draughts and carbon uptake capability close to that of the global average for forests (GPP and NEE mean values of 850 and 230 gC/yr respectively). We directly measured soil evaporation (E) using a modified soil respiration chamber (LI-COR) together with continuous eddy flux measurements of evapotranspiration (ET), heat-pulse measurements of tree transpiration (T) and precipitation (P). Soil water content (SWC) was recorded at three pits with TDRs horizontally installed to depth of 130 cm. Precipitation in the research area is characterized by sporadic rain events occurring during the winter months (Nov.-Apr.). Most rain-showers (82%) were smaller than 5 mm, enabling to increase superficial SWC and maintain high moisture level (above 20%) during most of the wet season. In contrast, SWC of the root zone (15-130 cm) depended on larger and less common rainstorms resulting with greater infiltration depths and showing high inter-annual variability. Annual E in this open canopy forest was 102±8 mm and accounted for approximately 44% of ET. T showed somewhat larger inter-annual variability (129±21 mm), clearly associated with storm pattern. Both seasonal and inter-annual trends of T were well synchronized with SWC at the root zone and with GPP. High water holding capacities of the soil resulted in increased SWC at depth only during mid-winter, which due to low E from these depths, was maintained in some cases throughout the dry summer, presumably increasing tree water uptake. Tree water use showed increased efficiency with increased storm P. Exceptional was an extremely dry year in which relatively high T was measured but not high GPP, presumably due to aboveground stress effect. The results indicated that except for extreme conditions, increased storm intensity could compensate for decreased precipitation at our region, contributing to the maintenance of relatively high productivity of the forest under drying conditions associated with global warming.
H13B-0924
Digital Image Analysis of Flowering in the Repeat-Blooming Creosotebush (Larrea tridentata) in Relation to Climatic Factors
Recent studies have shown that phenology (seasonal timing of life cycle events) is an effective integrator of the impacts of climate change on natural systems. Thus, understanding the climate signals that activate plant phenological responses (e.g., flowering and leaf production) will allow for improved modeling efforts and more effective ecosystem management strategies to mitigate the effects of climate change. With warmer and drier weather patterns predicted for the Desert Southwest in the coming century it can be expected that plant phenological patterns will be altered as a result. The most dominant and widespread shrub species of the warm desert ecosystems of North America is the creosotebush (Larrea tridentata). Consequently, creosotebush has a major impact on the structure, functioning, and flow of resources (i.e., carbon, water, and energy) in these regions, and when in bloom serves as an abundant and reliable food source for hundreds of pollinating insects that synchronize their emergence with flowering time. In this study, we hypothesize that frequency, duration and abundance of flowers in the repeat-blooming creosotebush are regulated by (1) temperature during the spring and (2) soil moisture below the depth of atmospheric demand in the summer. We make use of daily digital images from three stations at the Santa Rita Experimental Range in southern Arizona. These stations are located within the footprint of an eddy covariance tower, where continuous records of precipitation, air temperature, soil temperature, soil moisture at various depths, and net radiation are also being collected. Unlike more discrete methods used to observe seasonal changes in vegetation, use of daily images results in a continuous record that can be directly compared to micrometeorological data, allowing us to evaluate the bloom-up response of creosotebush alongside (1) air temperature, (2) soil temperature, and (3) soil water content fluctuations across time. We show that this technique increases data collection efficiency, enables better quantification of phenological events, and provides a continuous record of phenological activity that can be linked to climate records across time.