IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium, 2008
... These measurements included pigment contents, leaf optical properties, total leaf C:N ... mod... more ... These measurements included pigment contents, leaf optical properties, total leaf C:N ... model for noninvasive estimation of chlorophyll, carotenoids, and anthocyanin contents in ... instrument for passive remote sensing 1. Measurements of sunlight-induced chlorophyll fluorescence ...
Optical Sensors and Sensing Systems for Natural Resources and Food Safety and Quality, 2005
ABSTRACT Fluorescence of foliage in the laboratory has proven more rigorous than reflectance for ... more ABSTRACT Fluorescence of foliage in the laboratory has proven more rigorous than reflectance for correlation to plant physiology. Especially useful are emissions produced from two stable red and far-red chlorophyll fluorescence (ChlF) peaks centered at 685 nm and 735 nm. Methods have been developed elsewhere to extract steady state solar induced fluorescence (SIF) from apparent reflectance of vegetation canopies/landscapes using the Fraunhofer Line Depth (FLD) principal. Our study utilized these methods in conjunction with field-acquired high spectral resolution canopy reflectance spectra obtained in 2004 and 2005 over corn crops and small tree plots of three deciduous species (red maple, tulip poplar, sweet gum). Leaf level measurements were also made of foliage which included ChlF, photosynthesis, and leaf constituents (photosynthetic pigment, carbon (C), and nitrogen (N) contents). As part of ongoing experiments, measurements were made on N application plots within corn (280, 140, 70, and 0 kg N/ha) and tree (0, 37.5, 75, 112.5, 150 kg N /ha) sites at the USDA/Agriculture Research Service in Beltsville, MD. SIF intensities for ChlF were derived directly from canopy reflectance spectra in specific narrow- band regions associated with atmospheric oxygen absorption features centered at 688 and 760 nm. The red/far-red SIF ratio (SIFratio) derived from these field reflectance spectra successfully discriminated foliar pigment ratios altered by N application rates in both corn crops. This ratio was also positively correlated to the C/N ratio at leaf and canopy levels, for the available corn data (e.g., 2004). No consistent N treatment or species differences in SIF were detected in the tree foliage, but additional 2005 data are forthcoming. This study has relevance to future passive satellite remote sensing approaches to monitoring C dynamics from space.
IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium, 2008
In situ directional reflectance observations were taken on trees of three species and were used t... more In situ directional reflectance observations were taken on trees of three species and were used to validate the Four-Scale Linear Model for AnIsotropic Reflectance (FLAIR). FLAIR-simulated spectra were satisfactory in general, especially for the shaded and shaded/sunlit mixed portions of the canopy but were species dependent. Good agreements with field observations were found when utilizing these simulations to derive band
ABSTRACT Modeling directional reflectance in conjunction with in situ measurements provides an op... more ABSTRACT Modeling directional reflectance in conjunction with in situ measurements provides an opportunity to quantitatively examine vegetation responses expressed under a variety of viewing geometries and illumination conditions and to mprove our understanding of physiology related to carbon exchange between plants and the atmosphere. Recent studies have demonstrated that light use efficiency can be remotely acquired by utilizing Photochemical Reflectance Index to account for physiological responses of foliage exposed to different illumination conditions. In this study, BRDF was simulated with three radiative transfer models, SAILh, rowMCRM and the FLAIR, and compared with in situ measurements for validations. During the summers of 2007 and 2008, field campaigns were conducted at experimental tree plots and a corn field maintained by the USDA BARC. Hyperspectral measurements (~1 nm) were acquired for sectors where illumination conditions for foliage were either sunlit, shaded, or mixed sunlit/shaded, based on the relative azimuth angle between the observer and the sun. The shaded foliage was associated with the darkspot of the BRDF while the sunlit canopy is situated in the hotspot. These measurements were utilized for model input and for validation, using the original spectra and vegetation indices derived from them. The agreements between model simulations and in situ measurements varied for the models used and varied among canopy illumination sectors and species. Simulations from the FLAIR model showed satisfactory results, especially for the shaded portions. For the corn field, the best agreements were simulations from rowMCRM. Simulations from SAILh were better for the sunlit canopy while reflectance generated with rowMCRM showed better agreement for both sunlit and shaded partitions. For the FLAIR model, the simulations showed better results in the visible spectrum while errors in SAILh- and rowMCRM- simulated reflectance were relatively uniform in the visible and NIR region. Simulated PRI and NDVI also showed satisfactory results
A collaborative field campaign was undertaken in an experimental USDA cornfield in Beltsville, MD... more A collaborative field campaign was undertaken in an experimental USDA cornfield in Beltsville, MD to examine the temporal dynamics of spectral bio-indicators derived from canopy-level spectrometer observations and Earth Exploring One (EO-1) Hyperion satellite imaging spectrometer data. Intensive diurnal field-based campaigns were conducted weekly over six weeks in 2007 and over 14 weeks in 2008, across a transect within the
The physiological status of trees within forest ecosystems determines their proper functioning. A... more The physiological status of trees within forest ecosystems determines their proper functioning. A decline in the Norway spruce forests of the Krušné Hory Mts., Czech Republic, has been reported since the early 1950´s. It was attributed to the combination of severe atmospheric pollution and climatic conditions. Due to the mining activities and the burning of coal in the power plants in the area a strong gradient of acidic deposition leading from heavy (the eastern part) to light (the western part) developed during the 1970´s and 1980´s. Although the load of SO2 has significantly decreased since 1991, the full recovery of the damaged by the earlier deposition forests is a long term process. The physiological status of the Norway spruce forests has been assessing using the ground truth data (biochemical and spectroscopic data) as well as the two hyperspectral data sets acquired in 1998 (ASAS sensor - NASA Goddard Space Flight Center) and in 2013 (APEX sensor - developed by a Swiss-Belg...
ABSTRACT Fluxes of water and carbon into the atmosphere are critical components in order to monit... more ABSTRACT Fluxes of water and carbon into the atmosphere are critical components in order to monitor and predict climate change. Spatial heterogeneity and seasonal changes in vegetation contribute to ambiguities in regional and global CO2 and water cycle dynamics. Satellite remote sensing is essential for monitoring the spatial and temporal dynamics of various vegetation types for the purposes of determining carbon and water fluxes. Satellite data from the EO-1 Hyperion sensor was acquired for five Earth Observing Satellite (EOS) sites, Mongu (Zambia, Africa), Konza Prairie (Kansas, USA), Duke Forest (North Carolina, USA), Barrow (Alaska, USA) and Sevilleta (New Mexico, USA). Each EOS site represented a distinct vegetative ecosystem type; hardwood forest, grassland, evergreen forest, lichens, and shrubland/grassland respectively. Satellite data was atmospherically corrected using the Atmosphere CORrection Now (ACORN) model and subsequently, the spectral reflectance data was extracted in the vicinity of existing flux towers. The EO-1 Hyperion sensor proved advantageous because of its high and continuous spectral resolution (10 nm intervals from 355 to 2578 nm wavelengths). The high spectral resolution allowed us calculate biophysical indices based on specific wavelengths in the electromagnetic spectrum that are associated with alterations in foliar chemistry and plant membrane structure (i.e., vegetation stress) brought upon by many environmental factors. Previous studies have focused on relationships within a specific site or vegetation community. This study however, incorporated many sites with different vegetation types and various geographic locations throughout the world. Monitoring the fluctuations in vegetation stress with contemporaneous environmental conditions and carbon flux measurements from each site will provide better insight into water and carbon flux dynamics in many different biomes. Noticeable spectral signatures were identified based on site specific phenology, as well as distinct spectral shapes for each vegetation type. Preliminary results suggest a high correlation (r=0.77-0.86) between CO2 flux and various biophysical indices that were primarily associated with the red-edge inflection point (i.e., 670-780 nm wavelength), regardless of vegetation type. Soil heat flux measurements also correlated well with the biophysical indices. These findings demonstrate the advantages of high-resolution spectral imaging for Earth observations, as well as for monitoring regional and global water and carbon fluxes to understand their affects on climate change.
ABSTRACT Coastal tropical and subtropical environments, where there are distinct seasonal shifts ... more ABSTRACT Coastal tropical and subtropical environments, where there are distinct seasonal shifts in precipitation, can be highly susceptible to environmental changes caused by increasing anthropogenic pressure (e.g., urbanization, deforestation) in addition to natural "press and pulse" events, such as sea-level rise, tropical storms, and a changing climate. These man-made and natural perturbations directly affect the quality and quantity of water flowing through the ecosystem, both on the surface and subsurface. Changes in groundwater and surface water interactions will impact ecological communities, including highly vulnerable coastal mangrove communities. Nearly 1,445 km2 of mangroves cover Everglades National Park along the southern and southwestern coast of Florida. Rising sea levels, a predicted drier climate, and increased water demand may accelerate the landward migration of salt water intrusion which poses threats to the ecological communities along this coastal ecotone. This is a growing concern for the region and it is necessary that we understand the present hydrogeologic conditions to better monitor and model the future and inevitable changes to the coastal environment. The purpose of this preliminary study was to test the feasibility of measuring water quality indirectly from the spectral responses of mangrove vegetation on a regional scale. Spectra-derived biophysical indices were used to assess various relationships between the spectral signatures of the 3 main mangrove species (i.e., Avicennia germinans, Rhizophora mangle, and Laguncularia racemosa) and the ionic and nutrient concentrations in the porewater (i.e., 20cm and 100cm depths), surface water, and groundwater of the mangrove ecotone. Water samples from these sources were collected during the dry season, a transitional period, and the wet season at three sites in large, high-biomass mangroves along Shark River and two sites in dwarf, low-biomass, mangroves along Taylor River. Water samples were analyzed for major ions (e.g, Cl-, SO42-, Na2+, Mg2+, K+, and Ca2+) and nutrients (e.g., total organic carbon, N and P). The spectral responses of each of the mangrove species were collected in-situ within a few days of the water sampling. Initial results illustrate good correlations (R2>0.65; P<0.05) between various spectra-derived biophysical indices (e.g., EVI, NDVI) and porewater chloride concentrations. Other correlations demonstrate complex relationships between total N and P concentrations and site-specific mangrove spectra, suggesting physiological differences of nutrient uptake induced by salinity-related stress. The findings suggest the potential for upscaling these relationships using airborne and satellite hyperspectral imagery (e.g., AVIRIS, Hyperion) in order to monitor salt-water intrusion remotely on a regional scale. Further investigations with this research could provide insight to water and carbon flux dynamics within the Everglades and similar coastal mangrove ecosystems throughout the world.
IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium, 2008
... These measurements included pigment contents, leaf optical properties, total leaf C:N ... mod... more ... These measurements included pigment contents, leaf optical properties, total leaf C:N ... model for noninvasive estimation of chlorophyll, carotenoids, and anthocyanin contents in ... instrument for passive remote sensing 1. Measurements of sunlight-induced chlorophyll fluorescence ...
Optical Sensors and Sensing Systems for Natural Resources and Food Safety and Quality, 2005
ABSTRACT Fluorescence of foliage in the laboratory has proven more rigorous than reflectance for ... more ABSTRACT Fluorescence of foliage in the laboratory has proven more rigorous than reflectance for correlation to plant physiology. Especially useful are emissions produced from two stable red and far-red chlorophyll fluorescence (ChlF) peaks centered at 685 nm and 735 nm. Methods have been developed elsewhere to extract steady state solar induced fluorescence (SIF) from apparent reflectance of vegetation canopies/landscapes using the Fraunhofer Line Depth (FLD) principal. Our study utilized these methods in conjunction with field-acquired high spectral resolution canopy reflectance spectra obtained in 2004 and 2005 over corn crops and small tree plots of three deciduous species (red maple, tulip poplar, sweet gum). Leaf level measurements were also made of foliage which included ChlF, photosynthesis, and leaf constituents (photosynthetic pigment, carbon (C), and nitrogen (N) contents). As part of ongoing experiments, measurements were made on N application plots within corn (280, 140, 70, and 0 kg N/ha) and tree (0, 37.5, 75, 112.5, 150 kg N /ha) sites at the USDA/Agriculture Research Service in Beltsville, MD. SIF intensities for ChlF were derived directly from canopy reflectance spectra in specific narrow- band regions associated with atmospheric oxygen absorption features centered at 688 and 760 nm. The red/far-red SIF ratio (SIFratio) derived from these field reflectance spectra successfully discriminated foliar pigment ratios altered by N application rates in both corn crops. This ratio was also positively correlated to the C/N ratio at leaf and canopy levels, for the available corn data (e.g., 2004). No consistent N treatment or species differences in SIF were detected in the tree foliage, but additional 2005 data are forthcoming. This study has relevance to future passive satellite remote sensing approaches to monitoring C dynamics from space.
IGARSS 2008 - 2008 IEEE International Geoscience and Remote Sensing Symposium, 2008
In situ directional reflectance observations were taken on trees of three species and were used t... more In situ directional reflectance observations were taken on trees of three species and were used to validate the Four-Scale Linear Model for AnIsotropic Reflectance (FLAIR). FLAIR-simulated spectra were satisfactory in general, especially for the shaded and shaded/sunlit mixed portions of the canopy but were species dependent. Good agreements with field observations were found when utilizing these simulations to derive band
ABSTRACT Modeling directional reflectance in conjunction with in situ measurements provides an op... more ABSTRACT Modeling directional reflectance in conjunction with in situ measurements provides an opportunity to quantitatively examine vegetation responses expressed under a variety of viewing geometries and illumination conditions and to mprove our understanding of physiology related to carbon exchange between plants and the atmosphere. Recent studies have demonstrated that light use efficiency can be remotely acquired by utilizing Photochemical Reflectance Index to account for physiological responses of foliage exposed to different illumination conditions. In this study, BRDF was simulated with three radiative transfer models, SAILh, rowMCRM and the FLAIR, and compared with in situ measurements for validations. During the summers of 2007 and 2008, field campaigns were conducted at experimental tree plots and a corn field maintained by the USDA BARC. Hyperspectral measurements (~1 nm) were acquired for sectors where illumination conditions for foliage were either sunlit, shaded, or mixed sunlit/shaded, based on the relative azimuth angle between the observer and the sun. The shaded foliage was associated with the darkspot of the BRDF while the sunlit canopy is situated in the hotspot. These measurements were utilized for model input and for validation, using the original spectra and vegetation indices derived from them. The agreements between model simulations and in situ measurements varied for the models used and varied among canopy illumination sectors and species. Simulations from the FLAIR model showed satisfactory results, especially for the shaded portions. For the corn field, the best agreements were simulations from rowMCRM. Simulations from SAILh were better for the sunlit canopy while reflectance generated with rowMCRM showed better agreement for both sunlit and shaded partitions. For the FLAIR model, the simulations showed better results in the visible spectrum while errors in SAILh- and rowMCRM- simulated reflectance were relatively uniform in the visible and NIR region. Simulated PRI and NDVI also showed satisfactory results
A collaborative field campaign was undertaken in an experimental USDA cornfield in Beltsville, MD... more A collaborative field campaign was undertaken in an experimental USDA cornfield in Beltsville, MD to examine the temporal dynamics of spectral bio-indicators derived from canopy-level spectrometer observations and Earth Exploring One (EO-1) Hyperion satellite imaging spectrometer data. Intensive diurnal field-based campaigns were conducted weekly over six weeks in 2007 and over 14 weeks in 2008, across a transect within the
The physiological status of trees within forest ecosystems determines their proper functioning. A... more The physiological status of trees within forest ecosystems determines their proper functioning. A decline in the Norway spruce forests of the Krušné Hory Mts., Czech Republic, has been reported since the early 1950´s. It was attributed to the combination of severe atmospheric pollution and climatic conditions. Due to the mining activities and the burning of coal in the power plants in the area a strong gradient of acidic deposition leading from heavy (the eastern part) to light (the western part) developed during the 1970´s and 1980´s. Although the load of SO2 has significantly decreased since 1991, the full recovery of the damaged by the earlier deposition forests is a long term process. The physiological status of the Norway spruce forests has been assessing using the ground truth data (biochemical and spectroscopic data) as well as the two hyperspectral data sets acquired in 1998 (ASAS sensor - NASA Goddard Space Flight Center) and in 2013 (APEX sensor - developed by a Swiss-Belg...
ABSTRACT Fluxes of water and carbon into the atmosphere are critical components in order to monit... more ABSTRACT Fluxes of water and carbon into the atmosphere are critical components in order to monitor and predict climate change. Spatial heterogeneity and seasonal changes in vegetation contribute to ambiguities in regional and global CO2 and water cycle dynamics. Satellite remote sensing is essential for monitoring the spatial and temporal dynamics of various vegetation types for the purposes of determining carbon and water fluxes. Satellite data from the EO-1 Hyperion sensor was acquired for five Earth Observing Satellite (EOS) sites, Mongu (Zambia, Africa), Konza Prairie (Kansas, USA), Duke Forest (North Carolina, USA), Barrow (Alaska, USA) and Sevilleta (New Mexico, USA). Each EOS site represented a distinct vegetative ecosystem type; hardwood forest, grassland, evergreen forest, lichens, and shrubland/grassland respectively. Satellite data was atmospherically corrected using the Atmosphere CORrection Now (ACORN) model and subsequently, the spectral reflectance data was extracted in the vicinity of existing flux towers. The EO-1 Hyperion sensor proved advantageous because of its high and continuous spectral resolution (10 nm intervals from 355 to 2578 nm wavelengths). The high spectral resolution allowed us calculate biophysical indices based on specific wavelengths in the electromagnetic spectrum that are associated with alterations in foliar chemistry and plant membrane structure (i.e., vegetation stress) brought upon by many environmental factors. Previous studies have focused on relationships within a specific site or vegetation community. This study however, incorporated many sites with different vegetation types and various geographic locations throughout the world. Monitoring the fluctuations in vegetation stress with contemporaneous environmental conditions and carbon flux measurements from each site will provide better insight into water and carbon flux dynamics in many different biomes. Noticeable spectral signatures were identified based on site specific phenology, as well as distinct spectral shapes for each vegetation type. Preliminary results suggest a high correlation (r=0.77-0.86) between CO2 flux and various biophysical indices that were primarily associated with the red-edge inflection point (i.e., 670-780 nm wavelength), regardless of vegetation type. Soil heat flux measurements also correlated well with the biophysical indices. These findings demonstrate the advantages of high-resolution spectral imaging for Earth observations, as well as for monitoring regional and global water and carbon fluxes to understand their affects on climate change.
ABSTRACT Coastal tropical and subtropical environments, where there are distinct seasonal shifts ... more ABSTRACT Coastal tropical and subtropical environments, where there are distinct seasonal shifts in precipitation, can be highly susceptible to environmental changes caused by increasing anthropogenic pressure (e.g., urbanization, deforestation) in addition to natural "press and pulse" events, such as sea-level rise, tropical storms, and a changing climate. These man-made and natural perturbations directly affect the quality and quantity of water flowing through the ecosystem, both on the surface and subsurface. Changes in groundwater and surface water interactions will impact ecological communities, including highly vulnerable coastal mangrove communities. Nearly 1,445 km2 of mangroves cover Everglades National Park along the southern and southwestern coast of Florida. Rising sea levels, a predicted drier climate, and increased water demand may accelerate the landward migration of salt water intrusion which poses threats to the ecological communities along this coastal ecotone. This is a growing concern for the region and it is necessary that we understand the present hydrogeologic conditions to better monitor and model the future and inevitable changes to the coastal environment. The purpose of this preliminary study was to test the feasibility of measuring water quality indirectly from the spectral responses of mangrove vegetation on a regional scale. Spectra-derived biophysical indices were used to assess various relationships between the spectral signatures of the 3 main mangrove species (i.e., Avicennia germinans, Rhizophora mangle, and Laguncularia racemosa) and the ionic and nutrient concentrations in the porewater (i.e., 20cm and 100cm depths), surface water, and groundwater of the mangrove ecotone. Water samples from these sources were collected during the dry season, a transitional period, and the wet season at three sites in large, high-biomass mangroves along Shark River and two sites in dwarf, low-biomass, mangroves along Taylor River. Water samples were analyzed for major ions (e.g, Cl-, SO42-, Na2+, Mg2+, K+, and Ca2+) and nutrients (e.g., total organic carbon, N and P). The spectral responses of each of the mangrove species were collected in-situ within a few days of the water sampling. Initial results illustrate good correlations (R2>0.65; P<0.05) between various spectra-derived biophysical indices (e.g., EVI, NDVI) and porewater chloride concentrations. Other correlations demonstrate complex relationships between total N and P concentrations and site-specific mangrove spectra, suggesting physiological differences of nutrient uptake induced by salinity-related stress. The findings suggest the potential for upscaling these relationships using airborne and satellite hyperspectral imagery (e.g., AVIRIS, Hyperion) in order to monitor salt-water intrusion remotely on a regional scale. Further investigations with this research could provide insight to water and carbon flux dynamics within the Everglades and similar coastal mangrove ecosystems throughout the world.
Uploads
Papers by Petya Campbell