In the spring of 1986 the EOSAT Company and NASA Headquarters organized a workshop to consider: (... more In the spring of 1986 the EOSAT Company and NASA Headquarters organized a workshop to consider: (1) the potential value of space-acquired multiband thermal remote sensing in terrestrial research and commercial applications, and (2) the scientific and technological requirements for conducting such observations from the LANDSAT platform. The workshop defined the instrument characteristics of three types of sensors that would be needed to expand the use of thermal information for Earth observation and new commercial opportunities. The panels from two disciplines, geology and evapotranspiration/botany, along with the instrument panel, presented their recommendations to the workshop. The findings of these meetings are presented.
The most recent hydrological modeling studies have used satellite data to derive environmental dr... more The most recent hydrological modeling studies have used satellite data to derive environmental drivers of the models to augment or completely replace traditional meteorological station observations. We have begun to develop and validate existing algorithms for the MODIS (Moderate-resolution Imaging Spectroradiometer) sensor. MODIS data can be used to monitor various bio-physical parameters including surface temperature and air temperature that are
Thermal Remote Sensing in Land Surface Processing, 2004
There have been considerable advances in the estimation of land surface environmental conditions ... more There have been considerable advances in the estimation of land surface environmental conditions from satellite observations, particularly from thermal infrared remote sensing data (Running and Nemani 1988; Carlson et al. 1994; Norman et al. 1995; Prince and Goward ...
The relationship between the Normalized Difference Vegetation Index (NDVI) and surface skin tempe... more The relationship between the Normalized Difference Vegetation Index (NDVI) and surface skin temperature (Ts), the Temperature-Vegetation Index (TVX), generally exhibits a negative relationship where vegetation has a high NDVI and low temperature and bare soil has a low NDVI and higher temperature. This relationship has been utilized to study air temperature, soil moisture and surface moisture conditions as part of biospheric and hydrologic studies. A form of this relationship is part of the MOD 16 MODIS Evaporative Fraction (EF) product. We have found that the TVX relationship exhibits unusual behavior in some situations. We derived TVX slope and air temperature estimations using AVHRR data and compared the output to data from the Oklahoma Mesonet for the period of May 6 to August 20, 1999. This study shows that positive NDVI/Ts slopes were possibly caused by subpixel water bodies. In addition, NDVI/Ts slopes become increasingly negative after winter wheat harvest in July in Central...
... radiometer, for example mounted on a satellite) (Kumar and Monteith, 1982; Sellers, 1985; Asr... more ... radiometer, for example mounted on a satellite) (Kumar and Monteith, 1982; Sellers, 1985; Asrar et al., 1986; Sellers, 1987; Goward ... for plants having different biosynthetic pathways and respiration rates (Jarvis and Leverenz, 1983; Russell et al., 1989; Prince, 1991b; Runyon et ...
Improved processing algorithms are being developed, under the AVHRR Land Pathfinder II (ALP-II) P... more Improved processing algorithms are being developed, under the AVHRR Land Pathfinder II (ALP-II) Project, to process the entire Advanced Very High Resolution Radiometer (AVHRR) archive, including methods to produce surface temperature (T s ) measurements from the thermal infrared portion of the observation record. This new data set should assist in the monitoring of global terrestrial environmental conditions on seasonal to interannual time scales. The impact of both the calibration and filter function variations between National Oceanic and Atmospheric Administration (NOAA)-AVHRR sensors is addressed in this proposed ALP-II approach. Accounting for these sensor factors reduces the error of intersatellite comparison of surface temperatures by as much as 2.3 K. The environmental factors were also addressed, including the surface emissivity, large range of atmospheric water vapor (WV) across the globe, and the potential for large surfaceto-air temperature differences for land. To account for both the sensor and environmental factors in the ALP-II data set, we derive, from the MODTRAN-3 model, two sets of new coefficients, for low and high WV in the atmosphere. Two types of split-window equations were analyzed, the linear and quadratic forms of the land split-window equations. Our MODTRAN-based analysis suggests that the quadratic form is more accurate than the linear form, especially for high WV contents. These coefficients can be used in our LST algorithm in combination with a global emissivity map and WV data from Global Circulation Model (GCM).
measurements then~s.elves. Within the error,s of the methods used, estimation of standard nwteoro... more measurements then~s.elves. Within the error,s of the methods used, estimation of standard nwteorologieal shelter height air temperatures recorded at the time of satellite overpass appears possible. Further r@nenwnts of the remote sensing methods used here are possible and can be expected in the era of the National Aeronautics and Space Administration's' Earth Observing System.
In this study, four approaches to estimate atmospheric water vapor from Advanced Very High Resolu... more In this study, four approaches to estimate atmospheric water vapor from Advanced Very High Resolution Radiometer (AVHRR) observations were tested with data from the Boreal Ecosystem -Atmosphere Study (BOREAS) and the Oklahoma Mesonetwork. The approaches studied were (i) the split-window difference of the thermal channels by Dalu [Int. J. Remote Sens. 7 (1986) 1089.], (ii) the ratio of variances by Jedlovec [J. Appl. Meteorol. 29 (1990) 863.], (iii) the regression slope by Goward et al. [Ecol. Appl. 4 (1994) 322.], and (iv) a look-up table derived from radiative transfer model output. Although these techniques were primarily developed to estimate total column precipitable water, we used them to estimate near-surface water vapor, within a few meters of the surface. Near-surface water vapor is needed for hydrologic and biospheric modeling. Analysis showed the total column precipitable water to be highly correlated (r 2 = .79) with near-surface absolute humidity for clear-sky conditions at the BOREAS and the Oklahoma study sites. Correlation of all the retrieval techniques with ground observations was very low. For the split-window approach, water vapor can only be estimated on a per pixel basis and is ambiguous for anything but a single site. The regression slope and variance ratio techniques showed very little correlation with ground observations with r 2 = .02 when compared with data from BOREAS, and .17 for the variance ratio and .24 for the regression slope when compared with Mesonet data. The spatial variability of water vapor across the landscape hampers the use of these contextual approaches. The highest correlation was for the look-up table approach, with r 2 = .36 when compared with data from the BOREAS site. The look-up table was applied using AVHRR Channels 4 and 5 brightness temperatures, surface temperature, and near-surface air temperature. Surface temperature and air temperature were both estimated from the satellite readings. Combining the satellite data with air temperature measured at meteorological ground stations improved the correlation to .50. The relatively low r 2 values were at least partly due to spatial and temporal mismatches between surface and satellite measurements. Simulation of Moderate Resolution Imaging Spectrometer (MODIS) thermal Channels 29 (8.4 -8.7 mm), 31 (10.78 -11.28 mm), and 32 (11.77 -12.27 mm) brightness temperatures showed that Channels 31 and 32 provide similar information as AVHRR Channels 4 and 5. The additional thermal information provided by Channel 29 shows promise for future water vapor detection efforts.
Many recent studies have applied satellite remote sensing data to large-scale hydrologic and bios... more Many recent studies have applied satellite remote sensing data to large-scale hydrologic and biospheric modeling. It is widely accepted that the thermal infrared observations from the Advanced Very High Resolution Radiometer (AVHRR) have the potential to estimate land surface conditions, such as surface temperature, near surface air temperature, and near surface water vapor. In this study, algorithms to estimate all three variables are presented and applied to an area covering the state of Oklahoma for a six day period in August, 1994. The results were validated using ground observations from the 111 station Oklahoma Mesonet. Validation of the remote sensing algorithms with Mesonet observations produced comparable results to previous validation studies. In addition, the validation process revealed inadequacies in thermal modeling that had not been detected in previous validation studies leading to the development of a new approach to estimate atmospheric water vapor.
Photogrammetric Engineering & Remote Sensing, 2006
The Landsat-7 Enhanced Thematic Mapper Plus instrument has two selectable gains for each spectral... more The Landsat-7 Enhanced Thematic Mapper Plus instrument has two selectable gains for each spectral band. In the acquisition plan, the gains were initially set to maximize the entropy in each scene. One unintended consequence of this strategy was that, at times, dense vegetation saturated band 4 and deserts saturated all bands. A revised strategy, based on a land-cover classification and sun angle thresholds, reduced saturation, but resulted in gain changes occurring within the same scene on multiple overpasses. As the gain changes cause some loss of data and difficulties for some ground processing systems, a procedure was devised to shift the gain changes to the nearest predicted cloudy scenes. The results are still not totally satisfactory as gain changes still impact some scenes and saturation still occurs, particularly in ephemerally snow-covered regions. A primary conclusion of our experience with variable gain on Landsat-7 is that such an approach should not be employed on future global monitoring missions.
Abstract. Ecological modellers are preparing to address how the biosphere may respond to decadal-... more Abstract. Ecological modellers are preparing to address how the biosphere may respond to decadal-to century-scale changes in the Earth's environmental system. This endeavour has revealed a question for which no immediate answer is readily available: how do ...
Abstract Mapping of remotely-sensed data to a common projection can potentially degrade the quali... more Abstract Mapping of remotely-sensed data to a common projection can potentially degrade the quality of the data. The Canada Centre for Remote Sensing (CCRS) has developed the Geocoding and Compositing System (GEOCOMP) to map Advanced Very High Resolution Radiometer (AVHRR) data to a 1 km2 equal area projection. Using data from the Boreal Ecosystem-Atmosphere Study (BOREAS), we tested GEOCOMP's processing of individual pixels and its ability to preserve the spatial integrity of the original AVHRR observations. Our study shows that GEOCOMP processing does not significantly degrade the ability to derive environmental variables and indices using a contextual approach. Errors produced by the GEOCOMP system are largest for scenes with high radiometric contrast and the parts of scenes with large satellite view angles.
The crimes committed against humanity during World War II shocked the global community into actio... more The crimes committed against humanity during World War II shocked the global community into action. As the world continued to uncover the full extent of the Holocaust, international leaders in the newly formed United Nations (un) drafted the Universal Declaration of Human Rights and established a mandate to protect human rights and prevent atrocities. However, certain situations increase the likelihood of human-rights violations and therefore the need for human-rights monitoring of conflicts (Edwards and Koettl ). Because of ongoing abuses, the un, along with other governmental and nongovernmental organizations (ngos), conducts widespread and intensive human-rights monitoring campaigns. Broadly described by the Office of the High Commission for Human Rights as the "active collection, verification and immediate use of information to address human rights problems" (ohchr , ), human-rights monitoring documents any violations that include governmental and nonstate transgressions of human rights and the failures of the state to protect those rights. Remote sensing is a tool that improves detection of and potentially provides a deterrent to human-rights violations. Remote-sensing platforms include helicopters and fixed-wing or unmanned aerial vehicles. However, these platforms are limited in several aspects, including limited range and the need for basing facilities near conflict areas, which are often in very remote locations. These platforms also require overflight permission from the host government, which is often the alleged transgressor. For these reasons, organizations that monitor human rights have increasingly turned to space-based imagery to identify and document conflict. Organizations that use remote sensing are able to quickly order images of specific locations to verify an alleged human-rights violation, an activity known as "human-rights mapping." They are also able to use images to conduct a systematic human-rights monitoring campaign to observe a region. Information gathered using remote-sensing products such as DigitalGlobe imagery helps the un and ngos provide evidence of international law violations to fact-finding bodies and criminal courts and supports un peacekeeping missions (Pisano ). In the fu-
Due to climate differences, an extreme range in productivity occurs along a 250—km, west—east tra... more Due to climate differences, an extreme range in productivity occurs along a 250—km, west—east transect at °44° north latitude in western Oregon, USA, where coniferous evergreen forests dominate. As part of the Oregon Transect Ecosystem Research (OTTER) project, our objective was to evaluate how climate constrains net primary production (NPP) by limiting the utilization of intercepted photosynthetically active radiation (IPAR). The forests measured along the transect intercepted from 22% to 99.5% of the incident PAR. With data collected from recording meteorological stations installed near each site, we defined the hourly conditions when photosynthesis was partly or completely limited by drought, extreme humidity deficits, or frost. From this analysis we calculated that the fraction of incident PAR that could be utilized throughout the year ranged from 92% in the coastal rainforests to <25% in the juniper woodland. NPP varied from 3 to 26 Mg°ha—1°yr—1 with the fraction of belowgro...
The SAIL canopy radiative transfer model is employed to examine relations between the normalized ... more The SAIL canopy radiative transfer model is employed to examine relations between the normalized difference vegetation index (NDVI) and absorbed photosynthetically active radiation (APAR) in a herbaceous vegetation canopy. This analysis extends previous work by using a ...
In the spring of 1986 the EOSAT Company and NASA Headquarters organized a workshop to consider: (... more In the spring of 1986 the EOSAT Company and NASA Headquarters organized a workshop to consider: (1) the potential value of space-acquired multiband thermal remote sensing in terrestrial research and commercial applications, and (2) the scientific and technological requirements for conducting such observations from the LANDSAT platform. The workshop defined the instrument characteristics of three types of sensors that would be needed to expand the use of thermal information for Earth observation and new commercial opportunities. The panels from two disciplines, geology and evapotranspiration/botany, along with the instrument panel, presented their recommendations to the workshop. The findings of these meetings are presented.
The most recent hydrological modeling studies have used satellite data to derive environmental dr... more The most recent hydrological modeling studies have used satellite data to derive environmental drivers of the models to augment or completely replace traditional meteorological station observations. We have begun to develop and validate existing algorithms for the MODIS (Moderate-resolution Imaging Spectroradiometer) sensor. MODIS data can be used to monitor various bio-physical parameters including surface temperature and air temperature that are
Thermal Remote Sensing in Land Surface Processing, 2004
There have been considerable advances in the estimation of land surface environmental conditions ... more There have been considerable advances in the estimation of land surface environmental conditions from satellite observations, particularly from thermal infrared remote sensing data (Running and Nemani 1988; Carlson et al. 1994; Norman et al. 1995; Prince and Goward ...
The relationship between the Normalized Difference Vegetation Index (NDVI) and surface skin tempe... more The relationship between the Normalized Difference Vegetation Index (NDVI) and surface skin temperature (Ts), the Temperature-Vegetation Index (TVX), generally exhibits a negative relationship where vegetation has a high NDVI and low temperature and bare soil has a low NDVI and higher temperature. This relationship has been utilized to study air temperature, soil moisture and surface moisture conditions as part of biospheric and hydrologic studies. A form of this relationship is part of the MOD 16 MODIS Evaporative Fraction (EF) product. We have found that the TVX relationship exhibits unusual behavior in some situations. We derived TVX slope and air temperature estimations using AVHRR data and compared the output to data from the Oklahoma Mesonet for the period of May 6 to August 20, 1999. This study shows that positive NDVI/Ts slopes were possibly caused by subpixel water bodies. In addition, NDVI/Ts slopes become increasingly negative after winter wheat harvest in July in Central...
... radiometer, for example mounted on a satellite) (Kumar and Monteith, 1982; Sellers, 1985; Asr... more ... radiometer, for example mounted on a satellite) (Kumar and Monteith, 1982; Sellers, 1985; Asrar et al., 1986; Sellers, 1987; Goward ... for plants having different biosynthetic pathways and respiration rates (Jarvis and Leverenz, 1983; Russell et al., 1989; Prince, 1991b; Runyon et ...
Improved processing algorithms are being developed, under the AVHRR Land Pathfinder II (ALP-II) P... more Improved processing algorithms are being developed, under the AVHRR Land Pathfinder II (ALP-II) Project, to process the entire Advanced Very High Resolution Radiometer (AVHRR) archive, including methods to produce surface temperature (T s ) measurements from the thermal infrared portion of the observation record. This new data set should assist in the monitoring of global terrestrial environmental conditions on seasonal to interannual time scales. The impact of both the calibration and filter function variations between National Oceanic and Atmospheric Administration (NOAA)-AVHRR sensors is addressed in this proposed ALP-II approach. Accounting for these sensor factors reduces the error of intersatellite comparison of surface temperatures by as much as 2.3 K. The environmental factors were also addressed, including the surface emissivity, large range of atmospheric water vapor (WV) across the globe, and the potential for large surfaceto-air temperature differences for land. To account for both the sensor and environmental factors in the ALP-II data set, we derive, from the MODTRAN-3 model, two sets of new coefficients, for low and high WV in the atmosphere. Two types of split-window equations were analyzed, the linear and quadratic forms of the land split-window equations. Our MODTRAN-based analysis suggests that the quadratic form is more accurate than the linear form, especially for high WV contents. These coefficients can be used in our LST algorithm in combination with a global emissivity map and WV data from Global Circulation Model (GCM).
measurements then~s.elves. Within the error,s of the methods used, estimation of standard nwteoro... more measurements then~s.elves. Within the error,s of the methods used, estimation of standard nwteorologieal shelter height air temperatures recorded at the time of satellite overpass appears possible. Further r@nenwnts of the remote sensing methods used here are possible and can be expected in the era of the National Aeronautics and Space Administration's' Earth Observing System.
In this study, four approaches to estimate atmospheric water vapor from Advanced Very High Resolu... more In this study, four approaches to estimate atmospheric water vapor from Advanced Very High Resolution Radiometer (AVHRR) observations were tested with data from the Boreal Ecosystem -Atmosphere Study (BOREAS) and the Oklahoma Mesonetwork. The approaches studied were (i) the split-window difference of the thermal channels by Dalu [Int. J. Remote Sens. 7 (1986) 1089.], (ii) the ratio of variances by Jedlovec [J. Appl. Meteorol. 29 (1990) 863.], (iii) the regression slope by Goward et al. [Ecol. Appl. 4 (1994) 322.], and (iv) a look-up table derived from radiative transfer model output. Although these techniques were primarily developed to estimate total column precipitable water, we used them to estimate near-surface water vapor, within a few meters of the surface. Near-surface water vapor is needed for hydrologic and biospheric modeling. Analysis showed the total column precipitable water to be highly correlated (r 2 = .79) with near-surface absolute humidity for clear-sky conditions at the BOREAS and the Oklahoma study sites. Correlation of all the retrieval techniques with ground observations was very low. For the split-window approach, water vapor can only be estimated on a per pixel basis and is ambiguous for anything but a single site. The regression slope and variance ratio techniques showed very little correlation with ground observations with r 2 = .02 when compared with data from BOREAS, and .17 for the variance ratio and .24 for the regression slope when compared with Mesonet data. The spatial variability of water vapor across the landscape hampers the use of these contextual approaches. The highest correlation was for the look-up table approach, with r 2 = .36 when compared with data from the BOREAS site. The look-up table was applied using AVHRR Channels 4 and 5 brightness temperatures, surface temperature, and near-surface air temperature. Surface temperature and air temperature were both estimated from the satellite readings. Combining the satellite data with air temperature measured at meteorological ground stations improved the correlation to .50. The relatively low r 2 values were at least partly due to spatial and temporal mismatches between surface and satellite measurements. Simulation of Moderate Resolution Imaging Spectrometer (MODIS) thermal Channels 29 (8.4 -8.7 mm), 31 (10.78 -11.28 mm), and 32 (11.77 -12.27 mm) brightness temperatures showed that Channels 31 and 32 provide similar information as AVHRR Channels 4 and 5. The additional thermal information provided by Channel 29 shows promise for future water vapor detection efforts.
Many recent studies have applied satellite remote sensing data to large-scale hydrologic and bios... more Many recent studies have applied satellite remote sensing data to large-scale hydrologic and biospheric modeling. It is widely accepted that the thermal infrared observations from the Advanced Very High Resolution Radiometer (AVHRR) have the potential to estimate land surface conditions, such as surface temperature, near surface air temperature, and near surface water vapor. In this study, algorithms to estimate all three variables are presented and applied to an area covering the state of Oklahoma for a six day period in August, 1994. The results were validated using ground observations from the 111 station Oklahoma Mesonet. Validation of the remote sensing algorithms with Mesonet observations produced comparable results to previous validation studies. In addition, the validation process revealed inadequacies in thermal modeling that had not been detected in previous validation studies leading to the development of a new approach to estimate atmospheric water vapor.
Photogrammetric Engineering & Remote Sensing, 2006
The Landsat-7 Enhanced Thematic Mapper Plus instrument has two selectable gains for each spectral... more The Landsat-7 Enhanced Thematic Mapper Plus instrument has two selectable gains for each spectral band. In the acquisition plan, the gains were initially set to maximize the entropy in each scene. One unintended consequence of this strategy was that, at times, dense vegetation saturated band 4 and deserts saturated all bands. A revised strategy, based on a land-cover classification and sun angle thresholds, reduced saturation, but resulted in gain changes occurring within the same scene on multiple overpasses. As the gain changes cause some loss of data and difficulties for some ground processing systems, a procedure was devised to shift the gain changes to the nearest predicted cloudy scenes. The results are still not totally satisfactory as gain changes still impact some scenes and saturation still occurs, particularly in ephemerally snow-covered regions. A primary conclusion of our experience with variable gain on Landsat-7 is that such an approach should not be employed on future global monitoring missions.
Abstract. Ecological modellers are preparing to address how the biosphere may respond to decadal-... more Abstract. Ecological modellers are preparing to address how the biosphere may respond to decadal-to century-scale changes in the Earth's environmental system. This endeavour has revealed a question for which no immediate answer is readily available: how do ...
Abstract Mapping of remotely-sensed data to a common projection can potentially degrade the quali... more Abstract Mapping of remotely-sensed data to a common projection can potentially degrade the quality of the data. The Canada Centre for Remote Sensing (CCRS) has developed the Geocoding and Compositing System (GEOCOMP) to map Advanced Very High Resolution Radiometer (AVHRR) data to a 1 km2 equal area projection. Using data from the Boreal Ecosystem-Atmosphere Study (BOREAS), we tested GEOCOMP's processing of individual pixels and its ability to preserve the spatial integrity of the original AVHRR observations. Our study shows that GEOCOMP processing does not significantly degrade the ability to derive environmental variables and indices using a contextual approach. Errors produced by the GEOCOMP system are largest for scenes with high radiometric contrast and the parts of scenes with large satellite view angles.
The crimes committed against humanity during World War II shocked the global community into actio... more The crimes committed against humanity during World War II shocked the global community into action. As the world continued to uncover the full extent of the Holocaust, international leaders in the newly formed United Nations (un) drafted the Universal Declaration of Human Rights and established a mandate to protect human rights and prevent atrocities. However, certain situations increase the likelihood of human-rights violations and therefore the need for human-rights monitoring of conflicts (Edwards and Koettl ). Because of ongoing abuses, the un, along with other governmental and nongovernmental organizations (ngos), conducts widespread and intensive human-rights monitoring campaigns. Broadly described by the Office of the High Commission for Human Rights as the "active collection, verification and immediate use of information to address human rights problems" (ohchr , ), human-rights monitoring documents any violations that include governmental and nonstate transgressions of human rights and the failures of the state to protect those rights. Remote sensing is a tool that improves detection of and potentially provides a deterrent to human-rights violations. Remote-sensing platforms include helicopters and fixed-wing or unmanned aerial vehicles. However, these platforms are limited in several aspects, including limited range and the need for basing facilities near conflict areas, which are often in very remote locations. These platforms also require overflight permission from the host government, which is often the alleged transgressor. For these reasons, organizations that monitor human rights have increasingly turned to space-based imagery to identify and document conflict. Organizations that use remote sensing are able to quickly order images of specific locations to verify an alleged human-rights violation, an activity known as "human-rights mapping." They are also able to use images to conduct a systematic human-rights monitoring campaign to observe a region. Information gathered using remote-sensing products such as DigitalGlobe imagery helps the un and ngos provide evidence of international law violations to fact-finding bodies and criminal courts and supports un peacekeeping missions (Pisano ). In the fu-
Due to climate differences, an extreme range in productivity occurs along a 250—km, west—east tra... more Due to climate differences, an extreme range in productivity occurs along a 250—km, west—east transect at °44° north latitude in western Oregon, USA, where coniferous evergreen forests dominate. As part of the Oregon Transect Ecosystem Research (OTTER) project, our objective was to evaluate how climate constrains net primary production (NPP) by limiting the utilization of intercepted photosynthetically active radiation (IPAR). The forests measured along the transect intercepted from 22% to 99.5% of the incident PAR. With data collected from recording meteorological stations installed near each site, we defined the hourly conditions when photosynthesis was partly or completely limited by drought, extreme humidity deficits, or frost. From this analysis we calculated that the fraction of incident PAR that could be utilized throughout the year ranged from 92% in the coastal rainforests to <25% in the juniper woodland. NPP varied from 3 to 26 Mg°ha—1°yr—1 with the fraction of belowgro...
The SAIL canopy radiative transfer model is employed to examine relations between the normalized ... more The SAIL canopy radiative transfer model is employed to examine relations between the normalized difference vegetation index (NDVI) and absorbed photosynthetically active radiation (APAR) in a herbaceous vegetation canopy. This analysis extends previous work by using a ...
Uploads
Papers by Samuel Goward