Mission | People | Organization | Research | Education | Project | Data | Tools | Links
Remote Sensing of Plant Canopies
Mosquito Control and Coastal Hydro-Hypsometry
Earth Science & Public Health
Project Leader: Drs. Gilberto A. Vicente, (L) GMU and Nancy Maynard, NASA/GSFCBackground
One of the main objectives working for the NASA/GSFC Earth Science & Public Health Program is to bring land, atmosphere and ocean satellite data/products to users unfamiliar with satellite data/products in their original format, but knowledgeable in the GIS environment. The project consists in identifying satellite derived data/products relevant to the Health/Environment user community and implementing a multi-scale remotely sensed imagery delivery system readily available to GIS users from the county to the federal level and targeted to health related applications. This will be accomplished by 1), providing satellite data/products accessible to ArcGIS clients (ArcView, ArcEditor, ArcInfo) and 2), assisting in the establishment of local and specialized GIS database and ArcGIS servers based on the Remote Sensing Information Partnership (RSIP) concept.Technical Approach
Initial efforts have concentrated in the development of a collaboration project between the Goddard Distributed Active Archive Center (GDAAC), the Pennsylvania State Department of Environmental Protection (DEP) and Penn-State University, aimed to assist the Penn-State West Nile Virus (WNV) program, a concern of the Middle Atlantic region, as well.. Recent achievement include the development of an automated procedure to access EOS-HDF MODIS NDVI, Land Surface Temperature (LST) and Surface Reflectance (SR) and deliver these products in the GEO-TIF (GIS compatible) and JPEG formats through the Web to the Penn-State DEP users.The next year’s plan includes using the Penn-State WNV collaboration program as a model and proof-of-concept on how NASA remote sensing data can be put to operational use in health related applications. The system will be expanded to include other health relate problems such as Lyme disease, Asthma, and Cancer related to UV radiation exposure.
Milestones
Next 6 months: Use the Penn-State WNV collaboration program as a model and proof-of-concept on how NASA/GSFC remote sensing data can be put to operational use in health related applications. The program will focus in the search for the links between the Penn-State DEP fieldwork activities to combat mosquito breading sites, and the NASA satellite data and products. The goal is to assist county mosquito control staff to schedule their sampling and control work. The key to success will consists in helping the Penn-State DEP and Penn-State University in the development of a simple, dynamic, automated and descriptive model that ties the mosquito larvae life cycle development with NASA remote sensing derived surface temperature, near surface air temperature, evaporation rate, surface heights and terrain information and vegetation coverage.Next year: Expand the lessons learned from the NASA/GSFC participation in the Penn-State WNV program to include other health relate problems as Lyme disease, Asthma, Cancer related to UV radiation exposure, etc.
Remote Sensing of Plant Canopies
Atmospheric Nitrogen Deposition, Uptake of Contaminants, and Vegetation Stress
Project Leader: Dr. G. Taylor, GMUBackground
Monitoring of the spectral properties of plant canopies over varying spatial and temporal scales can provide valuable information in a host of issues in the environmental sciences. Three of these issues are addressed in this initiative.The first is a more accurate estimate of the flux of nitrogen (N) pollutants from the atmosphere to a watershed. For the Chesapeake Bay watershed, N inputs to the Bay are documented from runoff, but N contributions from the atmosphere are only crudely estimated. Dry deposition of odd-nitrogen gases (e.g., HNO3) is not measured. Estimating dry deposition of odd-N gases is typically constrained by the absence of landscape-specific leaf area indices (LAI) (surface for scavenging).
The second issue is one of contaminants in the upper soil horizons and the ability to monitor organics and inorganics using remote sensing. The soil-plant-continuum suggests that plant canopies may serve as a “window” on the contamination in the subtending soil. This initiative has value in the development of tools to solve problems related to soil contamination on public and private lands.
The third is plant stress. Plants are challenged by a variety of environmental factors. Some stresses are specific in their biochemical mode of action, whereas others disrupt general plant functions. It is proposed that some stressors have distinct signatures detectable using HSI.
The stakeholders for this project are state/federal agencies and NGO’s responsible for managing the Chesapeake as a natural resource.
Technical Approach
Dry N deposition will be estimated inferentially using the big leaf model that links air chemistry data (i.e., HNO3) with LAI by watershed type. The remote sensing data will be used to generate monthly LAI by landscape type (e.g., forests, wetland). The big leaf model will be used to integrate the two data sets, producing dry N deposition in units of kg N ha-1 month-1per landscape type.
The progress to date has focused on three steps. The first is the downloading and activation of the Big-Leaf Model to serve as the “coupler” linking the air quality data with the landscape-level estimates of LAI. The second is the identification of air quality data sets. The search has inventoried state level data from Maryland, Pennsylvania, Virginia and the District of Columbia. The value of the selected sites is being judged relative to QA/QC requirements and the percentage of completeness. The third activity has focused on the development of LAI for the watershed using remote sensing data. Greenness LAI examples have been developed and are shown in the figure below (Figure 8) for a peninsula in the Potomac River.
The technical approach for the study of contaminants will use a combination of field and laboratory studies and based on locations of known contamination. Plants will be grown in the lab under defined conditions with known levels of contaminant uptake. Hand-held HSI measurements will be made at intervals and the data compared with similar measurement on the same species under field conditions.
Progress to date has been limited to two activities. The first has been the selection of plant species. The second activity has been iterative discussions with state (Virginia) and federal (US DOD) agencies regarding the highest priority contaminants in the Chesapeake Watershed to use in the study.
The technical approach for the plant stress studies will be similar to that for contaminants. Using a whole-plant gas exchange chamber capable of simulating a variety of environmental conditions (e.g., temperature, ozone, drought), plants will be treated with single stress situations and their spectral properties measured in real time.
Initiation of this project will not commence until 2003.
THEMES
These initiatives are linked to a set of stakeholders in the public and private sector. Estimating dry deposition of N has value to land management organizations in the public and private sector, providing input to management decisions regarding source apportionment and the efficacy of control options. The initiative devoted to contaminant uptake has a near-term application in waste management needs. The final initiative is more exploratory. The stakeholders are the public and private landowners who manage terrestrial resources for productivity (e.g., agriculture) be beneficial in minimizing impacts on productivity and sustainability.Milestones
Estimate dry deposition of N for the Chesapeake WatershedFuture Milestones
Develop protocol for identifying spectral signatures for a select suite of organics and inorganic contaminants and plant species under laboratory and field conditions
Develop protocol for identifying spectral signature for a select suite of plant stressors under laboratory conditions
Mosquito Control and Coastal Hydro-Hypsometry
Project Leader: Dr. Thomas R. Allen (L) , Political Science and Geography, ODU
Dr. George F. Oertel, Ocean, Earth and Atmospheric Sciences, ODUThis plan of work provides an overview of the progress and planned continuation and augmentation of the VAccess grant at Old Dominion University. The project develops a paired-study focus on mosquito control users and coastal managers as end-users of remote sensing data and research analysis tools.
A. Remote Sensing in Mosquito Control and Disease Prevention: Continuing Research
Background:
The application of remote sensing to mosquito control has involved the development of a database integrating field and laboratory specimens of mosquitoes and associated serological analyses, GIS database development and assimilation, and a seasonal and multisensor time series image database. The overall approach has been to develop a prototype system for applying Landsat and other imagery to assist mosquito control. To date, the project has shown progress in spatial interpolation of mosquito trap locations in Fairfax County, Virginia, over a period of six months’ field data. The method developed and undergoing calibration applies geostatistics to imagery and GIS data for improved spatial interpolation of trap data. The results provide a more accurate regionalization of mosquito abundance and distributions in space and time in a GIS interface.Technical Approach: Surveillance Strategy
As a result of our modeling progress, a generic protocol (and algorithm for ArcGIS) are in preparation that will provide a tool for integrating imagery and GIS data for mapping habitats. This will subsequently be tested and calibrated in current study area of Henrico County, Virginia. Following this modeling calibration, the model will be validated in the City of Chesapeake, Virginia, through a cooperative agreement with the city’s large Washington Borough Mosquito Control District, adjacent to the Great Dismal Swamp. This site is selected for its potent arboviral reservoir community and vectors in proximity to the large human population of sou thisde Hampton Roads. Several cases of equine encephalitis and traps and birds containing West Nile Virus and associated mortality have recently been documented here.Our preliminary spatial and statistical analysis of imagery and geocoded mosquito traps are now in preparation for presentation and publication. Further work is required to calibrate and validate the method, to assay and quantify the utility of time series imagery, and to test for interannual variation and robustness of the modeling in Fairfax. The final tool to be developed is a series of risk surfaces, spatially calculated, using imagery and modeled human and vector behavior maps. This output will serve to improve the effectiveness of future surveillance and control, for public health officials and the cooperating private industry adopters. The target end-users include both public and private sector personnel, for whom a website and tutorial modules are planned.
B. Coastal Basins and Environmental Health: An Augmentation Activity for VAccess Year 2
Background:
Urbanization in the coastal zone continues apace with ongoing environmental changes, resulting in a grand challenge to managers to differentiate, monitor, and mitigate resource conflicts and environmental change. Sea level rise and natural disturbance events as well as coastal land use threaten the survival of wetlands and aquatic resources. With rising sea level, survival of wetlands and Submerged Aquatic Vegetation (SAV) is dependent on the availability of new surfaces for colonization. Lateral transgression or build-up with high rates of vertical sedimentation may produce these surfaces. The loss of SAV from the coastal bays has already had a great socio-economic effect on the local communities by essentially eliminating the scallop industry. In addition, wetlands, estuaries, tidal creeks and coastal bays often serve as receptacles for a variety of pollutants from anthropogenic activities.Technical Approach: Basin Hydro-Hypsometry and Flushing
If the rate of pollutant delivery is not buffered, or if the pollutants are not “flushed” from these environments, loading may cause severe damage to the environmental health. The economic consequences of coastal bay ecosystem alteration are often harsh, affecting local food supplies, the health of swimmers and consumers of seafood, local tourism, and aesthetic components which affect quality of life and economic prosperity. Remote sensing can dramatically augment traditional field assays and GIS databases across regions and regimes of coastal environments. Pollutant loading in aqueous environments is certainly a function of an often overlooked phenomenon, flushing characteristics of the basin. This factor requires knowledge of water level cycles (tides) and basin hypsography (the relationship between elevation and surface area). Remote sensing, with integrated GIS, provides a new avenue for mapping and modeling of this process in coastal estuarine environments.The key to applying a “hazard index” to land margins is based in hypsography and water-level change. The magnitudes of marsh loss, coastal flooding, vegetative buffers and flushing are directly related to these simple parameters. Our goal is to develop a protocol based in remote sensing and GIS for obtaining hypsograhic data in small watersheds. One product of the protocol might be a “flushing index” for small bays and estuaries that would provide managers with a risk-assessment tool for these environments. Others might be to assess the risk of marsh loss or coastal flooding.
Protocol Development
• Defining terrestrial boundaries and ‘mericana’ g sub-aqueous zones of small tidal watersheds
• Modeling sub-aqueous surfaces with bathymetry
• Estimation of flushing regimes based on hydraulic turn-over time
• Stratified Environmental Testing
• Developed high-intensity urban basin
• Rural agricultural basin
• Tidal regime sensitivity
• Outreach Application
Data Needs – VAccess, NASA, and Partner InstitutionsLandsat ETM+ and Ikonos digital imagery, tidal gauge data, digital bathymetric and topographic data, Radarsat and ENVISAT imagery for mosquito habitat assessment. Focus on Virginia (Fairfax, Richmond, Hampton Roads, and Eastern Shore)
Milestones
1. September 2002 to November 2002
Website inauguration for mosquito and coastal health projects, image acquisition and preprocessing, GIS data integration, tidal gauge data, equipment acquisition, equipment setup & installation, design protocols and stratification, change detection and time series spectral analysis of mosquito trap sites.2. December 2002 to February 2003
Implement and test protocols, remote sensing image processing, accuracy assessment, 2002 mosquito trap data analysis, risk index development for mosquito distributions.3. March 2003 to February 2003
Derive hydraulic turnover times, design index for coefficients of exchange, develop flushing index and “environmental health-risk” index, document best protocols, geostatistical modeling of mosquito data (2002) and model calibration.4. June 2003 to August 2003
Mosquito model validation and risk index application programming, Hampton Roads, Virginia. Submit results for publication, prepare outreach materials and tutorial module, present at professional conferences.
GIS Support
Dr. David Wong, GMUBackground:
GIS represents a cross-cutting technology that supports several key activities in VAccess – MAGIC. The specific activities currently supported are described below.
Technical Approach:
Meadowood Farm BLM Project (with Dr. George Taylor)
• a prototype procedure has already been developed to import field gathered data and GPS data into GIS (ArcView). But the procedure is not very efficient and requires significant human intervention and manual reformatting of data. A more automated procedure based upon shareware interfacing with the GPS unit and GIS will be tested to modify the existing procedure.
• So far, data of a small portion of the property have been processed and imported into GIS. The data for the rest of the site has yet to be gathered and processed. (See figure 11)
• Even though the data (ecological inventory) are associated with the sampled points on the map, the data have yet to be shown more effectively in maps to disseminate the information gathered. This stage requires some development in cartographic design and techniques.
Fairfax County Wetland Project (with Dr. George Taylor)
• Several data sets have been identified as potentially useful for this project. More detailed analyses of these data layers are required to see if they can help identify wetland areas.
• A statistical/overlay procedure conceptually similar, but operationally different from Cross Correlation Analysis has yet to be developed to detect wetland changes over time.
• Field data gathered from selected wetland sites based upon current surveys have to be imported into GIS so that they can be compared with historical wetland data (National Wetland Inventory and other images) and existing images.
Mosquito Spatial Surveillance Model (with Dr. Tom Allen)• Vulnerable population groups have been identified as children and elders. Their facility locations (child-care, elementary schools, and elder care facilities), associated with these population groups, have been captured in GIS data layers. These data will be used to select areas that more likely will require mosquito monitoring.
• Dr. Tom Allen is in the process of using various types of remotely sensed data and GIS data layers to characterize the landscape / land cover to identify areas potential mosquito breeding sites.
• We will obtain more mosquito monitoring data for Fairfax from the mosquito control company. These data will be imported into GIS to serve as the basis for estimating mosquito population or density. We will develop various interpolation / spatial statistical techniques in this estimation process in conjunction with landscape / land cover data.
• We plan to combine the estimates of mosquito population / density and the locations of the at-risk population to formulate an improved mosquito monitor scheme, eventually.
• Risk maps will be compiled to indicate the risk level of being exposed to mosquitoes.
• We expect to expand our project using Henico county and the City of Chesapeake to verify our model.Milestones:
2nd Quarter BLM GIS layers and selected maps
Wetlands field data in GIS system
Identify past and current data layers for wetland areas
Mosquitio population densities mapped for Fairfax County
Merger of populations at risk with mosquito density estimates4th Quarter BLM Data set, map products, GIS data layers
Test model for Henico County
Wetland alternate methods formulated
A paper was presented at Geoinformatics 2002, Nanjing, May 30 - June 3, 2002."Formulating a Spatial Monitoring Strategy for Mosquito Control: Identifying Vulnerable Population" by Tom Allen, ODU, and David Wong, GMU.
VAccess-MAGIC © 2001-2002. All rights reserved
