Impacts of spatial heterogeneity on the measurement and modeling of land-atmosphere interactions

Joseph Gerard Alfieri, Purdue University

Abstract

Surface characteristics, such as the type and spatial distribution of land cover, exert a critical control on both the turbulent exchange of mass and energy at the land-atmosphere interface and the subsequent meteorological, hydrological, and environmental processes. This research used a combination of in-situ, airborne, remotely sensed, and modeled data to address broad questions related to the role of land-atmosphere exchange processes in the earth’s climate system while focusing on the effects of fine-scale surface heterogeneity. Concentrating on the domain of the 2002 International H2O Project, this research was conducted in three phases. The first phase of the research sought to quantify the spatial variability in both the surface characteristics and airborne surface flux measurements. The former was achieved using spatial and categorical statistics historically associated with landscape ecology and natural resource management. It was found that while the types of land cover were similar, the proportions and degree of aggregation varied significantly across the domain. The analysis of the spatial variability of the flux measurements was achieved using statistical and process-based methods. Using the statistical methods, it was found that the variability in the turbulent fluxes varied significantly from flight-to-flight. Based on analyses using the process-based approach, the spatial variations in the turbulent fluxes were most closely tied to variations in surface properties impacting the available energy and, more specifically, the soil heat flux. The second phase of the research, linking the surface heterogeneity with the variability in the airborne flux measurements, was achieved using a combination of footprint, spatial statistical, and GIS analysis techniques. Both heuristic and observational studies showed that the contribution from any given area of interest (e.g. land use type) depended not only on the fraction of the footprint it represented, but also on its spatial distribution within the footprint. The final phase of the research applied the insights gained during the earlier phases of the research to improve the ability of the Noah land surface model, and the related modeling systems (e.g. WRF/Noah, HRLDAS, NLDAS), to represent surface heterogeneity. The core research associated with this phase of the research was the development of a novel evaluation scheme that overlays and weights the model output using the footprint of the observation data used for the validation process. It was found that by using the footprint to provide spatial context, the model could be better assessed. However, there is an important caveat, when the model data is aggregated to a coarse resolution or model runs are conducted at low resolutions, the benefit of using the footprint are lost. Other components of this phase of the research include an analysis of the Jarvis Scheme used by the model to estimate transpiration. Overall, this research underscores the importance of quantifying surface heterogeneity in order to correctly apply and interpret flux data, whether measured or modeled, when seeking to understand land-atmosphere exchange processes and their role in the climate system.

Degree

Ph.D.

Advisors

LeMone, Purdue University.

Subject Area

Hydrologic sciences|Atmospheric sciences|Remote sensing

COinS