Influence of land processes on weather and climate
Land-atmosphere coupling
Water resources in a changing climate
Ongoing Research
Contributing to the development of the next version of ACCESS for CMIP6
Reinventing subgrid tiling of the land and land-atmosphere coupling for weather and climate models
Implementing the CABLE model improvements in Noah and CLM
Quantifying the interplay between uncertainties in numerical solutions, parameters, and elected processes in soil moisture transport
Representing CABLE in the GSWP3 and SOIL-MIP modeling studies
Exploring benefits, limitations, and possible opportunities for optimizing LSM parameters
Past Research
CESM/CLM (Climate model/land model from NCAR)
Developed a new numerical solution to the highly nonlinear PDE governing soil moisture transport now used in CLM
Formulated an implicit aquifer parameterization providing simplest method of including soil-aquifer coupling in LSMs
Derived empirical relation between soil liquid,ice,temperature, and texture
WRF and LIS (Weather prediction model and land system)
Contributed to the development of new snow parameterizations for the Noah model
Implemented a simple irrigation model (prior to the release of official modules
Incorporated hydrological and flux parameterizations in CABLE-WRF using LIS
Developed ability to run land only simulations using land-atmosphere simulations
CABLE and ACCESS (Australian land and climate models)
Created conceptual model of subgrid scale soil moisture and runoff processes
Based on an assume subgrid PDF of soil moisture using a single parameter and subgrid topographic data
New parameterizations includes changes in both mean and variability (TOPMODEL approach in LSMs does not
Runoff and land-atmosphere fluxes formulated using subgrid properties
Added soil pore scale based physical processes to soil evaporation parameterizations
Unique among CMIP models due to physical basis, eliminating commonly used empirical methods
Reduced free model parameters, implemented without tuning
Fixed biases and limitations in CABLE
Brings simulated ET components in line with best estimates
Highlights necessity of including physical mechanisms in models
Demonstrated land-atmosphere feedbacks are of first order importance in forecasting irrigation demand
Quantified the inability of surface layer moisture satellite data to capture land-atmosphere coupling processes
Demonstrated that remote sensing based transpiration to provide drought-vegetation information beyond satellite greenness data
Showed that groundwater is essential for the differential drought response between Australian forests and grasslands
Provided first constraint on large scale transpiration anomalies with model-data fusions
Diagnosed why CABLE was unable to adequately simulate drought, and showed most LSMs do not reasonably simulate dry periods
Estimated changes in soil and groundwater separately using my modeling work and GRACE data in a data assimilation system
Quantified errors in near surface states and climate in several reanalysis products
Evaluated bias, correlation, and magnitude of variability in causing errors
Showed reanalysis products are most performant at monthly scales, with bias induced errors
Timescale below monthly have large errors due to correlation and incorrect variability
Created large scale estimates of how transpiration responds to precipitation anomalies
Compared satellite/model derived transpiration anomalies to multiple model dis-aggregated ET products and greenness indices
Simple greenness metrics from visible or microwave datasets similar to complex observations/model techniques
Extended common in situ observations of soil liquid and temperature to estimate soil ice contentGenerated separate estimate of soil and groundwater changes from GRACE data
Total Citation Count: 2037
Source: Google Scholar on Jan 18 2018
Selected research
Irrigation projections commonly use land only models (red points), neglecting land-atmosphere coupling, we showed that land-atmosphere feedbacks (blue) significantly change irrigation estimates. Land-atmo coupling moistens the near surface air (left panel), causing lower mean irrigation rates with increased variability (right panel).
Soil evaporation in weather/climate models is represented without regard to the dominant physics. A pore scale model, based on physics, greatly enhances model performance, removes the common ad hoc methods, and actually reduces uncertainty.
Soil moisture varies and runoff is generated at small spatial scales; below what can be included in weather/climate models. A new conceptual representation derived from the subgrid pdf was derived. They sound formulation results in subgrid dynamics in agreement with the limited data. Shown is a representation of the runoff, soil moisture, and groundwater processes.
