Progress 01/03/05 to 09/30/09
Outputs
Two studies concerning remote sensing for vegetation species identification were performed as part of this project, one involving a terrestrial invasive non-native plant, the other involving submerged aquatic vegetation. A study to use both 30-m resolution satellite imagery and meter-resolution satellite airborne imagery (NASA ITD) for mapping Brazilian peppertree (Schinus terebinthifolius Raddi) was performed. The 30-m resolution portion compared the effectiveness of biseasonal multispectral satellite imagery (Landsat-7) with single-season hyperspectral satellite imagery (EO-1) for the purpose of mapping peppertree stands. The meter-resolution portion compared the effectiveness of two currently-available types--multispectral satellite imagery (Digital Globe Quickbird) and hyperspectral airborne imagery (NASA ITD) for the purpose of mapping individual peppertrees within two relatively small sites (c. 8 km) in south Florida. This study comprised a comparison of
multispectral and hyperspectral approaches on a fair basis at the two commonly available resolutions (meter and 30 m). The imagery was processed from raw digital numbers into surface reflectance via calibration targets, and then into peppertree raster-GIS layers, and the results evaluated on the basis of extensive ground-truth obtained from differential-GPS assisted fieldwork. As part of the hyperspectral image processing, a spectral library of hyperspectral signatures for baseline pre-biocontrol peppertree and various other native and non-native plant species in the study area was produced. This work was funded by a grant from the Florida Dept. of Environmental Protection (FL-DEP). PI is J. D. Jordan of UF Ag&Bio Engineering Dept. Co-investigators included William A. Overholt of the Indian River Research and Education Center (IRREC) and James P. Cuda of the UF Entomology and Nematology Dept. The second study involved airborne hyperspectral imaging (ITRES CASI) for assessment of
submerged aquatic vegetation (SAV) in north Florida (lower St. Johns River basin). This second year of this two-year imaging project made use of technical recommendations from the first year's SAV study, involving refinement of the spectral bands selected and the use of in-situ calibration and evaluation targets. The imagery was processed from raw digital numbers into surface reflectance, to allow application hyperspectral techniques to produce a raster-GIS layer of SAV, and particularly certain natural communities, species, and conditions. As part of the image processing, a spectral library of hyperspectral signatures for these features (as well as non-SAV features such as mud, water, non-SAV vegetation, and algal blooms) was produced. This work was funded by the St. Johns River Water Management District (SJRWMD). PI is J. D. Jordan of UF Ag&Bio Engineering Dept. Cooperators included Courtney Hart, Dean Dobberfuhl, and others of the SJRWMD.
Impacts
Initial findings from the peppertree study indicate that at fine spatial resolution, airborne hyperspectral imagery (NASA ITD) produced a more accurate peppertree map than did satellite multispectral imagery (Digital Globe Quickbird). Further research comparing the peppertree-mapping results of single-date airborne hyperspectral imagery with those of multi-seasonal satellite multispectral imagery is recommended. The spectral library produced will be of ongoing applicability in future hyperspectral peppertree work by the PI and other remote sensing researchers. Once the biocontrol agents are released, the hyperspectral signatures of affected peppertrees will be subject to change, rendering the pre-biocontrol baseline invaluable for monitoring the effectiveness of the peppertree biocontrol program. The submerged aquatic vegetation airborne hyperspectral imaging study will be carried-on by the SJRWMD.
Publications
- Hart, C. H., D. Dobberfuhl, and J. D. Jordan, 2006. Using Hyperspectral Imagery to Detect Submerged Aquatic Vegetation. Proceedings, American Water Resources Association (AWRA) Specialty Conference "GIS and Water Resources IV", Houston, TX, 8-10 May 2006.
- Jordan, J. D., 2006. Hyperspectral Signature Baseline for Monitoring the Biocontrol of Schinus terebinthifolius Raddi in Florida. Final Report to the Florida Dept. of Environmental Protection, Tallahassee, FL, 14 June 2006.
- Jordan, Jonathan D., William A. Overholt, James P. Cuda, and Keng L. Huang, 2007. Baseline Map and Spectral Library of Pre-Biocontrol Schinus terebinthifolius Raddi in Florida. Florida Exotic Pest Plant Council (FLEPPC) 22nd Annual Symposium, Cocoa Beach, FL, April 30-May 3, 2007.
|
Progress 10/01/05 to 09/30/06
Outputs
Two studies concerning remote sensing for vegetation species identification were performed during this period as part of this project, one involving a terrestrial invasive non-native plant, the other involving submerged aquatic vegetation. In addition, a new study was planned involving remote sensing for a specific crop disease. A study to use high-resolution satellite (Digital Globe Quickbird) and hyperspectral airborne imagery (NASA ITD) for mapping Brazilian peppertree (Schinus terebinthifolius Raddi) was performed at two relatively small sites (c. 8 km) in south Florida, in order to compare multispectral and hyperspectral approaches on a fair basis, and to extract a hyperspectral signature baseline for pre-biocontrol peppertree. The imagery was processed from raw digital numbers into surface reflectance via in-situ calibration targets, and then into peppertree raster-GIS layers, and the results evaluated on the basis of extensive ground-truth obtained from
differential-GPS assisted fieldwork. As part of the hyperspectral image processing, a spectral library of hyperspectral signatures for baseline pre-biocontrol peppertree and various other native and non-native plant species in the study area was produced. This work was funded by a grant from the Florida Dept. of Environmental Protection (FL-DEP). PI is J. D. Jordan of UF Ag&Bio Engineering Dept. Co-investigators included William A. Overholt of the Indian River Research and Education Center (IRREC) and James P. Cuda of the UF Entomology and Nematology Dept. An airborne hyperspectral imaging (ITRES CASI) mission to study submerged aquatic vegetation (SAV) in north Florida (lower St. Johns River basin) was begun and is ongoing. This imaging made use of technical recommendations from the previous year's SAV study, involving refinement of the spectral bands selected and the use of in-situ calibration and evaluation targets. The imagery is currently being processed from raw digital numbers
into surface reflectance, with the intention of applying hyperspectral techniques to produce a raster-GIS layer of SAV, and particularly certain natural communities, species, and conditions. As part of the image processing, it is anticipated that a spectral library of hyperspectral signatures for some or all of these features (as well as non-SAV features such as mud, water, non-SAV vegetation, and algal blooms) may be produced. A portion of this work was funded by the St. Johns River Water Management District (SJRWMD); support for the remainder is being sought from SJRWMD and FL-DEP. PI is J. D. Jordan of UF Ag&Bio Engineering Dept. Cooperators included Courtney Hart, Dean Dobberfuhl, and others of the SJRWMD. A new study was planned involving the use of airborne hyperspectral imaging to assess Huanglongbing (HLB) disease in Florida citrus orchards. An initial proposal was made to the USDA Tropical Subtropical Agriculture Research (T-STAR) program. PI is W. Lee of UF Ag&Bio
Engineering Dept. Co-investigators include J. D. Jordan of the UF Ag&Bio Engineering Dept., and R. Ehsani, G. Albrigo, and J. C. V. Rodrigues of the Citrus Research and Education Center (CREC).
Impacts
Initial findings from the peppertree study indicate that at fine spatial resolution, airborne hyperspectral imagery (NASA ITD) produced a more accurate peppertree map than did satellite multispectral imagery (Digital Globe Quickbird). Further research comparing the peppertree-mapping results of single-date airborne hyperspectral imagery with those of multi-seasonal satellite multispectral imagery is recommended. The spectral library produced will be of ongoing applicability in future hyperspectral peppertree work by the PI and other remote sensing researchers. Once the biocontrol agents are released, the hyperspectral signatures of affected peppertrees will be subject to change, rendering the pre-biocontrol baseline invaluable for monitoring the effectiveness of the peppertree biocontrol program. A presentation at the Florida Exotic Pest Plant Council (FLEPPC) conference in 2007 and a refereed article are in-preparation. The submerged aquatic vegetation airborne
hyperspectral imaging study is ongoing. The portion of the work funded by the SJRWMD to-date covers only the initial calibration; support for the remainder is being sought from SJRWMD and FL-DEP. Support for the citrus HLB disease airborne hyperspectral imaging study is being sought from USDA. It is anticipated that this will be a multi-year study involving substantial fieldwork and labwork.
Publications
- Min, Min, Won Suk Lee, Thomas F. Burks, Jonathan D. Jordan, Arnold W. Schumann, and John K. Schueller, 2006. Design of a Hyperspectral Nitrogen Sensing System for Citrus. Computers and Electronics in Agriculture (submitted).
- Cornejo, Camilo, Dorota Z. Haman, Ramon L. Espinel, and Jonathan D. Jordan, 2006. Irrigation Potential of the TRASVASE System (Santa Elena Peninsula, Guayas, Ecuador). Journal of Irrigation and Drainage Engineering, 132(5): 453-462.
- Hart, Courtney, Dean Dobberfuhl, and Jonathan D. Jordan, 2006. Using Hyperspectral Imagery to Detect Submerged Aquatic Vegetation. American Water Resources Association (AWRA) Spring Specialty Conference GIS and Water Resources IV, Houston, TX, 8-10 May.
- Jordan, J. D., 2006. Hyperspectral Signature Baseline for Monitoring the Biocontrol of Schinus terebinthifolius Raddi in Florida. Final Report to the Florida Dept. of Environmental Protection, Tallahassee, FL, 14 June 2006.
- Lee, Won Suk, Thomas F. Burks, Jonathan D. Jordan, John K. Schueller, Masoud Salyani, Arnold W. Schumann, Francis J. Pierce, Eileen Perry, Joan R. Davenport, and Robert G. Stevens, 2005. Maintaining the Competitiveness of Tree Fruit Production through Precision Agriculture. Final Report to USDA, 6 Dec. 2005.
|
Progress 10/01/04 to 09/30/05
Outputs
Three studies were performed this period as part of this project, two concerning remote sensing for vegetation species identification, and one concerning remote sensing for vegetation nutrient level response. A study to use satellite imagery for mapping Brazilian peppertree (Schinus terebinthifolius Raddi) was performed to cover the state-wide distribution of this plant at a spatial resolution of 30 meters using Landsat-7 ETM+ imagery; a smaller pilot study area was covered at the same spatial resolution by EO-1 Hyperion imagery. The imagery was processed into a peppertree raster-GIS layer, based on U.S. Geological Survey orthophotos (spatial correction), ancillary data (spectral calibration), and fieldwork transects (classification training sites, and accuracy evaluation sites, for peppertree and non-peppertree vegetation). This work was funded by a combination of grants from the Florida Dept. of Environmental Protection (FL-DEP), the University of Florida School of
Natural Resources and Environment (SNRE), and the University of Florida Invasive Plants Working Group (IPWG). Co-investigators included William A. Overholt of the Indian River Research and Education Center (IRREC), James P. Cuda of the UF Entomology and Nematology Dept., and graduate student K. Liang Huang of UF Agricultural and Biological Engineering Dept. Airborne imagery from the CASI instrument was the basis for a remote sensing study of submerged aquatic vegetation (SAV) in the lower St. Johns River basin. The narrowband imagery was processed into raster-GIS layers of SAV and various non-SAV features (other plants, mud, algal blooms, water), based on fieldwork which included SAV transects conducted by the St. Johns River Water Management District (SJRWMD), and spectral reflectance measurements performed by the PI from a SJRWMD boat. A methodology was developed in this pilot study for utilizing the processed imagery and ground measurements to prepare spectral libraries and perform
hyperspectral feature classifications (via spectral-angle-mapper technique). This work was funded by the SJRWMD. Cooperators included Courtney Hart, Dean Dobberfuhl, Nadine Trahan, David Girardin, and others of the SJRWMD. A pilot study to assess the impacts of phosphorus levels on bahia-grass was performed using airborne multispectral imagery from the ADS-40 instrument. The ultra-fine spatial resolution imagery was obtained to cover the UF-IFAS Pine Acres facility near Citra, FL. The imagery was processed into raster-GIS layers of spectral reflectance and vegetation indices, based on fieldwork data (spectral calibration, spatial accuracy checks via professionally surveyed benchmarks) and laboratory analysis (grass foliage samples). The study site included experimental grass plots having various levels of fertilization. This work was supported by in-kind contribution (imagery) from 3001 Inc. of Gainesville, FL. Cooperators included John C. Craig of 3001 Inc., and Ismail Bogrekci,
Jasmeet Judge, and Orlando Lanni of the UF Agricultural and Biological Engineering Dept.
Impacts
The raster-GIS layer of statewide peppertree extent, which was produced from the peppertree study, is in the process of being converted to vector-GIS format for potential inclusion in the Florida Natural Area Inventory (FNAI) GIS database. Initial findings from this study indicated that single-season hyperspectral imagery (100+ narrow spectral bands) appeared to yield better results than multiple-season multispectral imagery (c. 10 wide spectral bands); however, this was based on satellite imagery of 30 meter spatial resolution (Landsat-7 ETM+ and EO-1 Hyperion). Further research comparing these two remote sensing approaches at finer spatial resolution was recommended. This research, to be performed by the same investigators using the finer resolution Quickbird satellite multispectral imagery together with airborne hyperspectral imagery, was funded by the Florida Dept. of Environmental Protection for 2005. From the SAV study, a methodology for spectral-library
extraction and hyperspectral feature classification from CASI airborne imagery was developed. Technical recommendations for SJRWMD's upcoming 2006-2007 SAV remote sensing missions were included in the final report. The bahia-grass phosphorus study produced technical recommendations for upcoming remote sensing vegetation nutrient studies planned by the UF Precision Agriculture and Remote Sensing (PARS) group. In addition, a detailed base-map of the Pine Acres facility has been produced, which has been shared with other UF researchers having field experiments at that location.
Publications
- Bogrekci, Ismail, Wonsuk Lee, and Jonathan D. Jordan, 2005. Airborne Hyperspectral Imaging for Sensing Phosphorus Concentration in the Lake Okeechobee Drainage Basin. Paper No. 5795-13, Proceedings, Society for Photo-optical Instrumentation Engineering (SPIE) Defense and Security Symposium, Orlando, FL, 29-31 March 2005.
- Bogrekci, Ismail, Wonsuk Lee, and Jonathan D. Jordan, 2005. Multispectral Image Analysis for Phosphorus Measurement in Bahia Grass. American Society of Agricultural Engineers (ASAE) Annual International Meeting, Tampa, FL, 17-20 July 2005.
- Cornejo, Camilo, Dorota Z. Haman, Ramon L. Espinel, and Jonathan D. Jordan, 2005 Use of an Evapotranspiration Model and a Geographic Information System (GIS) to Estimate the Irrigation Potential of the TRASVASE System in the Santa Elena Peninsula, Guayas, Ecuador. Trans. Am. Soc. Civil Engineers (submitted).
- Hart, Courtney, Dean Dobberfuhl, and Jonathan D. Jordan, 2005. Using Hyperspectral Imagery to Detect Submerged Aquatic Vegetation in the St. Johns River. Seven Hills Regional User Group (SHRUG) GIS Workshop, 7-9 Nov. 2005, Tallahassee, FL.
- Jordan, J. D., 2005. Spectral Processing Methodology Development for Airborne Imagery of Submerged Aquatic Vegetation in the Lower St. Johns River. Final report to St. Johns River Water Management District, Palatka, FL, 28 Sept. 2005.
- Jordan, Jonathan D., William A. Overholt, and James P. Cuda, 2005. Baseline Mapping via Remote Sensing for Monitoring the Biocontrol of Schinus terebinthifolius Raddi in Florida. DEP-IFAS Research Review, Gainesville, FL, 28-29 March 2005.
- Jordan, Jonathan D., Keng L. Huang, William A. Overholt, and James P. Cuda, 2005. Baseline Mapping via Remote Sensing for Monitoring the Biocontrol of Schinus terebinthifolius Raddi in Florida. Final Report to Florida Dept. of Environmental Protection, Tallahassee, FL, 1 June 2005.
|
|