Jean-Claude Thomas, Dan Sechrist
Pielke, Sr., Roger A., Steyaert, Louis T.
Developed algorithms for extrapolation of in-situ evapotranspiration measurements using statistical summaries of TM data
Produced map of evapotranspiration with l00m resolution for south Florida for the image date of 3/21/96
Tested and eliminated the possibility of transferring TM developed statistical techniques to AVHRR for improved temporal resolution
Developed spectra for samples of cattail, sawgrass, periphyton, and open water through in situ measurements
Developed co-registered, georeferenced data sets from TM, SPOT, AVHRR, STATSGO SOILS, and climate stations in GIS format
Calibration of TM and AVHRR data sets to radiance, reflectance, and apparent surface temperature
Development of procedures for statistical sampling and analysis of any georeferenced data set using a combination of GIS, image processing, and advanced statistical software
Fieldwork for this effort has included the collection of high-resolution reflectance spectra for a great number of vegetation and land surfaces. Also, vegetation biomass and other structural characteristics have been sampled at intensive field study sites. Along with other ground data such as water level, elevation, and land cover type, these data are being used to test the efficacy of data fields and vegetation maps derived from the remotely sensed data. Data from numerous airborne and satellite imaging systems have been georeferenced and pre-processed to facilitate data fusion and analysis. Databases of different temporal and spatial solutions (depending on extent) that depict changes in vegetation amount and vigor (through vegetation indexes) have been developed for small areas like the Everglades Nutrient Removal project area and the entire South Florida region. A vegetation map of the Southern Inland and Coastal Systems (SICS) model study area has been developed for the application of spatially distributed fields of vegetation flow resistance. A similar map is currently being produced for the Tides and Inflows to Mangroves of the Everglades (TIME) study area. Data from several different remote-sensing systems and in situ data collections have been fused for the development of other map products to include vegetation density, surface reflectance, and inundated areas, as well as the development of visually enhanced satellite image maps. Finally, spatial analysis of derived variables has been undertaken to address issues of scale important in aggregation for hydrodynamic modeling.
Satellite Image Mapping in the Big Cypress area
This task will produce a 1:100K satellite image map of Big Cyrpress area that will abut the two previous satellite image maps created through this project (i.e., The Southern Everglades and Northern Everglades image maps).
The image fusion and other cartographic procedures developed through this research project will be applied using additional data acquired for the region of the Big Cypress preserve. Procedures that produce tonal and resolution qualities that match those of previous image maps will be used so that one mosaic can be made of all the data for the region of South Florida below Lake Okeechobee. The image maps previously created have been widely used as an outreach and planning tool. The development of the map for the Big Cypress region is a logical conclusion to pre-restoration image map production. The requirement to match previous satellite image map characteristics makes the near-term execution of this task critical.
1) Link high-resolution remote sensed indices of vegetation characteristics with point-based measurements of vegetation characteristics. This will be accomplished using previously collected vegetation and remotely sensed data using multiple regression techniques.
2) Develop relationships between high-resolution remotely sensed vegetation indices and satellite-based (coarser resolution) vegetation indices.
3) Use spatial analysis to extrapolate vegetation index models throughout the TIME model domain using multi-date satellite imagery.
4) Populate hydrodynamic models with spatially distributed, multidate flow resistance indices based on the extrapolated vegetation parameters.
5) Evaluate model performance with and without fields of vegetation flow resistance.
Work in FY 2003 will build upon previous efforts by Jones to spatially extrapolate ET values measured at point locations through Edward German’s ET project. Research on data calibration and atmospheric correction has been undertaken for this task and a set of calibrated/atmospherically corrected satellite data has been generated. Jones and Sechrist will expand this data set so that more rigorous models can be developed and evaluated. While modeling efforts will focus on key subareas within the Everglades, data for locations in the Florida ET Network (Sumner) will also be leveraged for model development. In conjunction with the vegetation characterization activities, we will use ground-based measurements of vegetation density and biomass as calibration and validation data for remotely sensed estimates of vegetation characteristics that likely influence ET. Evapotranspiration maps (in GIS and hardcopy formats) will be produced for use in indexing ET in South Florida hydrologic models. Results will be evaluated on the basis of technical review, assessment using withheld ground data, and output impacts on hydrologic model performance.
Throughout the year, requests and opportunities arise for pilot studies to investigate the use of novel remote sensing and geospatial analysis techniques to gather information of importance to CERP objectives, water quality and flow modeling, ecological modeling, and even as an aid to other remote sensing efforts. Two specific pilot mapping activities currently planned for FY 2003 and FY 2004. The first is focused on the characterization of solution holes in the Rocky Glades. Solution holes in the that region may constitute critical refugia and other habitat. Little is known about their spatial distribution or structural characteristics. It is also not clear how water resource manipulation will impact the function of these holes. The objective of this subtask is to investigate the potential of remote sensing techniques for solution hole survey, characterization, and monitoring.
A second subtask is focused on periphytoon detection and mapping. Periphyton affects water flow, mercury methelation, and the reflectance recorded by remotely sensed imagery. Previous research has demonstrated that periphyton mapping may be possible using hyperspectral imaging techniques that are currently used operationally. The minimum objective of this pilot study is to determine whether the presence or absence of periphyton can be estimated through the use of operational remote sensing systems. A "perihyton index" is the goal. Our ability to conduct more sophisticated periphyton mapping research will be dependent on available data and collaborator resources. I f appropriate ground and remote sensing data are available, this work may be extended to included periphyton composition mapping.
In FY 2003, airborne imagery will be collected and analyzed for its efficacy in mapping the location and surface characteristics of solution holes. Airborne aerial photography will be evaluated through visual interpretation and compared against validation data collected in the field. In addition, bathymetric LIDAR data will be collected for examination in FY 2004.
1. Land surface characterization for hydrological and ecological modeling
We will complete the collection of multi-temporal leaf area index (LAI) measurements at various points within the Greater Everglades region. Methods of extrapolating LAI values from points to the region will be developed and tested. We will then use spatial analysis to characterize the spatial structure in LAI at multiple scales and use that characterization to develop and test techniques for assigning flow resistance coefficients that are adjusted for sub-cell heterogeneity to TIME model cells.
2. Greater Everglades focused Status and Trends Topical Report
Following USGS publication guidelines, we will compile and publish a USGS circular-like document using both reprinted and custom-generated papers. At present, we anticipate including the following: 1) Document introduction and overview of Everglades environmental issues/the concerted Everglades restoration project 2) Everglades vegetation history from sediment core pollen analysis 3) Modeling Everglades surface hydrodynamics 'getting the water right' 4) The impact of anthropogenic Twentieth Century land use change on sea breeze generated convective rainfall and sensible weather over the South Florida Peninsula 5) Hurricanes impacts on Everglades mangroves 6) A sampling framework for Everglades landcover change assessment 7] A sidebar regarding Satellite image maps as research, monitoring, and educational outreach tools
Field/remote sensing technique development for scaling studies, data calibration, and targeted CERP-MAP activities 1. Solution hole mapping pilot study: Using extremely high resolution orthographic imagery generated by post-processing the data we collect using our airborne imaging system, we will map the location and density of various types of solution holes for pilot study areas in the Rocky Glades region. This image data will be coordinated with field data collection on solution-hole content and characteristics. This activity is directly responsive to the performance measure (GE-A4) information need identified in the MAP (section 3.1.4.7 titled the Role of Solution Holes as Aquatic Refuges for Marsh Fishes and Other Aquatic Animals in Karst Wetlands) that calls for a "Pilot Study of Remote Sensing/Surveying Methods for Estimating Refuge Characteristics". Three work activities are being specifically addressed through this activity: a. Conduct a pilot study to test alternative remote sensing methods to determine their resolution and accuracy in estimating hole density, areas, and depths in the rocky glades. b. Validate the methods by comparison to results from standard land surveying methods. c. Determine optimum study designs and survey methods to characterize the density, areas, and depth distributions of solution holes in the rocky glades in a spatially explicit manner. 2. Cooperative, structured field experiment on periphyton hyperspectral analysis: Previous research by the principle investigator has established methods of differentiating gross differences in periphyton composition along environmental gradients using hyperspectral airborne imaging. While this experiment relied in part on field-collected handheld radiometry, detailed quantitative analyses of periphyton content was not possible because resources for detailed taxonomic analysis of the periphyton were unavailable. This year, in collaboration with the South Florida Water Management District, we will collect spectra over ground samples of periphyton that will then be carefully collected and analyzed using established SFWMD protocols and analytical resources. In this way, we will test whether spectral features identified through previous research are truly diagnostic of periphyton assemblage composition. This activity is directly responsive to trophic systems monitoring requirements for periphyton production, cover, and composition associated with the key uncertainty of vegetation mapping technology development (MAP Section 3.1.4.5) and vegetation mapping: "Using hyperspectral systems as a cost-effective way of mapping Everglades landscape and water quality patterns".
Construction of well-calibrated, high quality multi-resolution and multi-temporal databases for landscape-scale modeling and targeted CERP-MAP work activities.
This task is focused on the development and testing of methods for multi-temporal satellite data radiometric calibration and atmospheric correction to provide for most accurate and consistent land cover change analysis, biophysical remote sensing, and CERP monitoring. The objective is to build a remote sensed database that is:
1. Well-calibrated (converted to physical values with some mitigation of atmospheric effects) 2. Multi-scale (temporally: from event based to frequent; spatially: from point-based to regional) 3. Multi-spectral (panchromatic, hyperspectral, RADAR, LIDAR, etc.) 4. Extensively documented (metadata traces all processing actions).
Three different calibration and atmospheric correction algorithms will be implemented and rigorously evaluated for their efficiency and effectiveness in producing consistent, regional temporal series of satellite data for Everglades research and monitoring. This evaluation will be completed using the rich, previously assembled data base of Landsat TM, Landsat MSS, SPOT XS, and AVHRR data augmented with new acquisitions of MODIS, ASTER, Hyperion, and other satellite/airborne data. Because coverage by these sensor systems is regional and the ultimate use of these data is land surface change monitoring, this Task directly supports most restoration projects south of Lake Okeechobee.
Landscape dynamics for landscape model development and enhancement
This year, the study will begin testing hypotheses in three subject areas:
1. Landscape ecology: This activity will focus on a specific ecological premise regarding wetland/vegetation landscape pattern and extent (MAP sections 3.1.2.X) about pattern and directionality in Everglades wetland landscapes. Landscape metrics will be applied to study-derived multi-scale field and remote sensed data to understand the scale lengths and directions over which contemporary vegetation density varies in the Greater Everglades. Such analyses will also inform the development of higher resolution hydrologic models - another need identified in the DOISP (pg 81). 2. Change Detection: We will quantify the thresholds of land surface spectral change that are detectable in calibrated satellite data. Change detection techniques (e.g., image differencing and multi-temporal principle components analysis) will be applied to the calibrated satellite image library developed in Task 2 to determine the types of changes that can be detected and the timescale(s) over which changes occur. Because the Altantic Coastal Ridge will also be included in change detection analyses this year, information will be provided that supports the need to understand links between hydrology and ecology for the Biscayne Bay (DOISP pgs 66/67). 3. Vegetation/environment relationships: Multivariate visualization and statistical analyses will be applied to the Everglades Vegetation Database and High Accuracy Elevation Dataset to examine relationships among vegetation and topography that have been suggested through field-based research as documented in the literature. Greater understanding of vegetation/topography relationships will aid modeling and planning for habitat and species recovery projects (DOISP section 4).
Construction of well-calibrated, high quality multi-resolution and multi-temporal databases for landscape-scale modeling and targeted CERP-MAP work activities.
Three different calibration and atmospheric correction algorithms will be implemented and rigorously evaluated for their efficiency and effectiveness in producing consistent, regional temporal series of satellite data for Everglades research and monitoring. This evaluation will be completed using the rich, previously assembled data base of Landsat TM, Landsat MSS, SPOT XS, and AVHRR data augmented with new acquisitions of MODIS, ASTER, Hyperion, and other satellite/airborne data. This year we will use the LNWR as a focus area for satellite data calibration and correction accuracy assessment. The LNWR area includes numerous structures that afford assessment of geometric corrections applied to the satellite data. Additionally, the LNWR region includes several relatively spectrally invariant land surfaces for calibration and testing of atmospheric correction approaches. However, because coverage by these sensor systems is regional and the ultimate use of these data is land surface change monitoring, LNWR-focused research also directly supports most restoration projects south of Lake Okeechobee.
Change detection technique development using Loxahatchee information needs
Exploratory and structured experiments will be conducted to determine the amounts of change in LNWR land surfaces that can be experimentally and operationally detected. Change detection techniques (e.g., image differencing and multi-temporal principle components analysis) will be applied to the calibrated satellite image library developed in Task 1 to determine the types of changes that can be detected and the timescale(s) over which changes occur. Rather than prescribe the changes being targeted, the PI will look for changes in the imagery and then label those changes based on ancillary information. This is an empirical process in which the thresholds of change that are identifiable in the imagery will be determined and then compared against features documented by previous field surveys, high-resolution aerial photography, and current project field-work. Some tonal changes may be easy to identify (e.g., vegetation to open water or the opposite). Others, such as sawgrass to cattail or brush to sawgrass for example, will be more difficult to discern. Once the PI has determined what changes can be reliably detected and identified using our techniques and available imagery, additional funding will be pursued for multi-decade, comprehensive LNWR change identification.
EDEN Grid and Everglades elevation model development
This task provides the ground elevation data QA/QC and advanced digital elevation modeling required by the Everglades Depth Estimation Network (EDEN) and associated ecological monitoring activities. Task objectives include the development of high quality, region-wide elevation data bases, characterization of LIDAR and surveyor collected elevation data quality, and intelligent modeling of Everglades ground elevations given a variety of input data types and sources.
Three primary activities are envisioned for this task.
1. Development of the EDEN grid with multiple thematic attributes (e.g., elevation, elevation estimation confidence/quality, vegetation composition, etc.). 2. QA/QC and conflation of Airborne Height Finder, ground (professionally) surveyed, and LIDAR data. 3. Development of enhanced digital elevation models for the Greater Everglades Region.
Southern Everglades Satellite Image Map
U.S. Department of the Interior, U.S. Geological Survey, Center for
Coastal Geology
Comments and suggestions? Contact: Heather
Henkel - Webmaster
Generated by mp version 2.8.18 on Tue Jan 30 15:40:46 2007