H31K-01
What We Have Learned From Field Experiments In The 20 Years Since FIFE
The field campaigns of the First ISLSCP Field Experiment took place in Kansas in 1987 and 1989, and the field phase of BOREAS lasted from 1993 through 1996, after which a consortium of Canadian researchers took over and continued with the experimental infrastructure. Twenty years after FIFE, it is useful to assess what was attempted, what was done and what has been learned. In the early 1980's, climate modelers were making their first efforts at modeling the exchanges of energy, moisture and momentum between the lower atmosphere and the land surface within general circulation models ( GCMs). It was clear at that time, ca 1984, that the parameterizations in use were very unrealistic. Also, there was no global method to initialize or validate these parameterizations. Accordingly, the experimental goals of FIFE were to: (i) Collect the data required to develop and improve land surface-atmosphere parameterizations, and (ii) Develop satellite-based methods to initialize and validate these parameterizations on regional and global scales. The methodology used in FIFE involved taking measurements over a nested series of spatial scales using surface, airborne and satellite platforms above a grassland field site in Kansas. Designing and executing the experiment was complex as a variety of scientific disciplines and skills were required: airborne and surface flux measurement; soil physics, hydrology and plant physiology; and surface, airborne and satellite remote sensing. All of these observations had to be integrated with each other. The results were combined within a common data management system which was made available to the participating field researchers and the wider modeling community. BOREAS extended this approach to include studies of large-scale carbon fluxes and climate change within the Canadian boreal forest. What has been learned? Results from the experiments were used to greatly improve land surface parameterizations that have since been tested and incorporated within global weather and climate models, and new methods have been developed for interpreting satellite data into useful biophysical parameters for these models. Both of these developments will be essential to future studies of global change. This paper will summarize the highlights.
H31K-02
Building on BOREAS: Biosphere-Atmosphere Interactions and the Fluxnet-Canada Continental-Scale Transect
Fluxnet-Canada is a national-level research network that was built on the scientific and methodological advances developed during BOREAS and then applied them to a continental scale. The focus of Fluxnet-Canada is on understanding the role of climate and disturbance on the carbon cycle of Canadian forests and peatlands. Taking advantage of the infrastructure installed during BOREAS, we established an east-west transect of eddy covariance flux towers that encompasses many of Canada's important ecoregions from New Brunswick to British Columbia. Fluxnet-Canada, and its follow-on project, the Canadian Carbon Program, make continuous, multi- year, tower-based measurements of the net exchanges of carbon (C), water, and heat from forest and peatland ecosystems for approximately 30 sites. These measurements allow us to (a) examine the relationship between the inter-annual variability of C fluxes and climate, (b) examine the role of disturbance, (c) analyze the contribution of different ecosystem components to the net C flux; (d) explore the relationship between productivity and net ecosystem C exchange; and (e) evaluate ecosystem and climate models. This information is used to extrapolate the knowledge obtained from the specific flux sites in both space and time. A land surface flux network is essential to the development of the scientific concepts, tools, data bases, models, and decision support capabilities required to build an integrated carbon observation and prediction system for Canada and North America. BOREAS provided much of the initial scientific foundation for this effort. Some of the major results from the last five years of research will be presented, e.g., the influence of time since disturbance was found to be much greater generally than was the influence of current levels of interannual climate variability. The Canadian Carbon Program will further extend the scientific heritage of BOREAS by moving terrestrial carbon cycle science in Canada to its next phase of development. A major focus will be exploring the synergies of combining high- precision greenhouse gas concentration measurements with on-going measurements of ecosystem flux.
H31K-03
The Impact of FIFE on the Business of Science
The impact of FIFE on many different scientific disciplines is well documented. The manner in which FIFE was conducted contributed greatly to interdisciplinary research and with its Integrative Science Team assured that the primary questions addressed by FIFE were addressed. The legacy of FIFE also includes an impact on how the business of science is conducted by interdisciplinary, inter-agency, and international organizations. Follow-on field campaigns conducted for FED, BOREAS, HAPEX-Sahel, KUREX, OTTER, SMEX and others have all benefited from the pioneering organization, planning, execution, communication and follow-on activities that stem from FIFE-87. Lessons-learned and examples of their impact are presented.
H31K-04 INVITED
Advances in the Analytic Description of Canopy Radiation for Climate Models
Absorption of solar radiation drives canopy exchanges of energy, moisture, and carbon. Prior to FIFE Dickinson and Sellers advanced one dimensional analytic models for determining this absorption in the contexts of a climate model. Their models now form the basis for what is currently used in climate models. However, under many circumstances, the assumptions of such models about the geometry of canopies are too inaccurate for use without further improvement. They can be adequately constrained by remote sensing observations such as from MODIS to provide reasonable estimations of total absorbed solar energy. However, 3D geometric effects are important for determining the partitioning of this absorbed energy between different surfaces, e.g. canopy versus snow. More realistic 3D modeling of canopy radiation has advanced for remote sensing applications through use numerical models and lookup tables. Radiation measurements supporting such models have been made in multiple field programs. This talk will move from this historical perspective to a summary of what is now possible through use of analytic mathematics to characterize canopy radiation with 3D geometries. Such characterization provides several benefits: it is efficient enough to be useful for climate models and hence can evaluate the terrestrial radiative feedbacks of interest in that context; it describes the directionally of the outgoing radiation and so can be used in a climate model for assimilation of remote sensing information about the terrestrial system; its analytic structure provides insight into the underlying physics. It also is of relevance to determining of emissivity for vegetated surfaces in a climate model.
H31K-05
Consistent Partitioning of Solar Fluxes in the Soil-Vegetation-Atmosphere System from Satellite Observations: State-of-the-Art 20 Years after FIFE
We present and evaluate results from the application of an inversion method conducted using both MODIS and MISR derived broadband visible and near-infrared surface albedo products available during a full seasonal cycle over the US Konza prairie in Kansas. This method enables us to assimilate operational remote-sensing flux products into a state-of-the-art two-stream radiation transfer scheme suitable for Global Climate Models. The occurrence of snow during the winter and spring seasons is based on the analysis of the MODIS snow-products, the assimilation of which by our package translates into an adaptation of the prior values, both the maximum likelihood and width of the 2-D probability density functions (PDF), characterizing the soil background conditions of the vegetation canopy. Our results illustrate the capability of the inversion package to assess a meaningful partitioning of the solar fluxes between the soil, vegetation and atmosphere layers, as well as, the retrieval of associated two-stream model parameters (such as the effective LAI and the albedo of the vegetation background) along the year despite the rather high variability in the input products. Results from applications conducted over additional sites characterized by evergreen and deciduous forest canopy systems and instrumented following the FIFE design will be discussed as well.
H31K-06
Linking Fluxes with Remote Sensing in Field Studies
From the beginning in FIFE, flux measurements were key data sets and the sites with flux towers were important anchors in the field campaigns. A key goal of FIFE was to develop connections between remotely sensed observations and fluxes of water vapor and carbon, mostly through providing parameters for models. In the twenty years since FIFE there has been significant changes in both flux measurements and remote sensing technology. Eddy covariance systems for measuring ecosystem carbon and water fluxes have become more reliable and cheaper, resulting in more flux data collected in more locations and over longer time periods. Eddy covariance systems provide temporally rich (but spatially poor) measurements of carbon and water vapor flux at selected locations. On the other hand, satellite-based remote sensing provides a synoptic view of the world, yet has difficulty providing temporally continuous data at a scale that can be directly related to flux measurements. There are a number of issues that must be explored to develop robust links between remote sensed observations and fluxes. The SpecNet (Spectral Network) group of investigators has been formed to address these issues. Many of its members have a background with field experiments. Key issues include both temporal and spatial scaling. Both flux measurements and spectral reflectance measurements can be collected at different spatial scales: leaf, plant, patches, stands, on up to entire landscapes. The SpecNet investigators also have an interest in temporal scaling, collecting both flux and spectral reflectance measurements simultaneously over various time intervals; from minutes, hours, days, growing seasons, to multiple years. To collect these types of data, new field methods and technology are being developed. A particular interest of SpecNet investigators is the detection of changes in reflectance in specific spectral bands that are caused by biochemical changes in leaves as they respond to stresses and downregulating photosynthesis. Changes in apparent reflectance have been observed at the leaf level due to changes in the xanthophyll cycle and also due to solar-induced fluorescence. These changes are subtle, but by tightly linking both optical and gas exchange measurements in space and time, they can be detected at patch on up to stand scales. This presentation will review some of the fieldwork that has emerged to examine the remote sensing of water and carbon fluxes. http://specnet.info/
H31K-07
Multi-Angle Remote Sensing of Forest Light Use Efficiency
Tower flux measurements over forests continue to contribute valuable understanding to Earth Systems Science and lay the foundations for new remote sensing measurement techniques. Using analytic methods and tower- based, quasi-continuous, full-spectrum observations of a Douglas fir forest over nearly 6 months, we demonstrate that stress-induced variations in measured photochemical reflectance index (PRI) with canopy light use efficience (LUE) are a direct result of xanthophyll-induced changes in leaf level reflectance at 531 nm, and can be observed at multi-canopy scales These conclusions are drawn inferentially from both analysis and observation. The key to our findings is the ability to continuously measure the reflectance of the Douglas fir forest as a function of shadow fraction from the hotspot to the "cold spot" and a new finding that the magnitude of an normalized difference reflectance index (NDRI) formed from two spectral bands cannot vary significantly with shadow fraction unless the reflectance of the shaded and sunlit leaves differ in at least one of the reflectance bands. Our findings suggest a new sensor and methodology for the direct retrieval from space of xanthophyll-induced changes in the LUE by measuring PRI as a function of shadow fraction using a multi-angle spectrometer simultaneously retrieving both shadow fraction and PRI. Such observations, when used in combination with physiological models should provide a greatly improved capability to monitor surface atmosphere exchanges of carbon, water and heat.