Thomas J. Schmugge

RESEARCH ARTICLE: Using Unmanned Aerial Vehicles for Rangelands: Current Applications and Future Potentials

High resolution aerial photographs have important rangeland applications, such as monitoring vegetation change, developing grazing strategies, determining rangeland health, and assessing remediation treatment effectiveness. Acquisition of high resolution images by Unmanned Aerial Vehicles (UAVs) has certain advantages over piloted aircraft missions, including lower cost, improved safety, flexibility in mission planning, and closer proximity to the target. Different levels of remote sensing data can be combined to provide more comprehensive information: 15–30 m resolution imaging from space-borne sensors for determining uniform landscape units; < 1 m satellite or aircraft data to assess the pattern of ecological states in an area of interest; 5 cm UAV images to measure gap and patch sizes as well as percent bare soil and vegetation ground cover; and < 1 cm ground-based boom photography for ground truth or reference data. Two parallel tracks of investigation are necessary: one that emphasizes the utilization of the most technically advanced sensors for research, and a second that emphasizes the minimization of costs and the maximization of simplicity for monitoring purposes. We envision that in the future, resource management agencies, rangeland consultants, and private land managers should be able to use small, lightweight UAVs to satisfy their needs for acquiring improved data at a reasonable cost, and for making appropriate management decisions.

Modeling and Inversion in Thermal Infrared Remote Sensing over Vegetated Land Surfaces

Thermal Infra Red (TIR) Remote sensing allow spatializing various land surface temperatures: ensemble brightness, radiometric and aerodynamic temperatures, soil and vegetation temperatures optionally sunlit and shaded, and canopy temperature profile. These are of interest for monitoring vegetated land surface processes: heat and mass exchanges, soil respiration and vegetation physiological activity. TIR remote sensors collect information according to spectral, directional, temporal and spatial dimensions. Inferring temperatures from measurements relies on developing and inverting modeling tools. Simple radiative transfer equations directly link measurements and variables of interest, and can be analytically inverted. Simulation models allow linking radiative regime to measurements. They require indirect inversions by minimizing differences between simulations and observations, or by calibrating simple equations and inductive learning methods. In both cases, inversion consists of solving an ill posed problem, with several parameters to be constrained from few information. Brightness and radiometric temperatures have been inferred by inverting simulation models and simple radiative transfer equations, designed for atmosphere and land surfaces. Obtained accuracies suggest refining the use of spectral and temporal information, rather than innovative approaches. Forthcoming challenge is recovering more elaborated temperatures. Soil and vegetation components can replace aerodynamic temperature, which retrieval seems almost impossible. They can be inferred using multiangular measurements, via simple radiative transfer equations previously parameterized from simulation models. Retrieving sunlit and shaded components or canopy temperature profile requires inverting simulation models. Then, additional difficulties are the influence of thermal regime, and the limitations of spaceborne observations which have to be along track due to the temperature fluctuations. Finally, forefront investigations focus on adequately using TIR information with various spatial resolutions and temporal samplings, to monitor the considered processes with adequate spatial and temporal scales. 10.1 Introduction Using TIR remote sensing for environmental issues have been investigated the last three decades. This is motivated by the potential of the spatialized information for documenting the considered processes within and between the Earth system components: cryosphere [1–2], atmosphere [3–6], oceans [7–9], and land surfaces [10]. For the latter, TIR remote sensing is used to monitor forested areas [11–14], urban areas [15–17], and vegetated areas. We focus here on vegetated areas, natural and cultivated. The monitored processes are related to climatology, meteorology, hydrology and agronomy: (1) radiation, heat and water transfers at the soil–vegetation–atmosphere interface [18–24]; (2) interactions between land surface and atmospheric boundary layer [25]; (3) vegetation physiological processes such as transpiration and water consumption, photosynthetic activity and CO2 uptake, vegetation growth and

Comparison of Thermal Infrared Emissivities Retrieved With the Two-Lid Box and the TES Methods With Laboratory Spectra

Knowledge of surface emissivity in the thermal infrared (TIR) region is critical for determining the land surface temperature (LST) from remote-sensing measurements. If emissivity is not well determined, it can cause a significant systematic error in obtaining the LST. The main aim of this paper is to compare different methods for measuring accurate land surface emissivity in the field, namely, the box method and the temperature and emissivity separation (TES) algorithm. Field emissivities were compared with soil spectra from laboratory measurements. Emissivities were measured for the bands of a multispectral radiometer CE312-2 with effective wavelengths at 8.4, 8.7, 9.1, 10.6, and 11.3 mum, similar to the Advanced Spaceborne Thermal Emission and Reflection Radiometer TIR bands, and a wide channel 8-13 mum. The measurements were made at two sites in New Mexico: the White Sands National Monument and an open shrub land in the Jornada Experimental Range. The measurements show that for both sites the emissivities derived with the Box method agree with those derived with the TES algorithm for the 10.6 and 11.3 mum bands. However, the emissivities for the shorter wavelength bands are higher when derived with the Box method than those with the TES algorithm, with differences ranging from 2% to 7%. The field emissivities agree within 2% with the laboratory spectrum for the 8-13-, 11.3-, and 10.6-mum bands. However, the field and laboratory measurements in general differ from 2.4% to 9% for the shorter wavelength bands, with the larger value most likely caused by variations in soil moisture.

Validation of Emissivity Estimates from ASTER and MODIS Data

Two distinctly different approaches are used to extract emissivity information from ASTER and MODIS data. ASTER uses an intuitive empirical relationship between the range of emissivities in the 5 ASTER bands and their minimum value. With its greater swath MODIS is able to uses the day / night pair of observations to obtain the emissivities. The combination of the two approaches should provide robust estimation of the land surface emissivity.

Analysis of ASTER Emissivity Product Over an Arid Area in Southern New Mexico, USA

The accuracy of thermal infrared emissivities derived from Advanced Spaceborne Thermal Emission and Reflectance radiometer (ASTER) was assessed in an arid area in southern New Mexico, which includes the White Sands National Monument (WSNM) during 2006-2008. ASTER emissivities retrieved by the temperature and emissivity separation (TES) algorithm were directly compared with laboratory measurements of samples from WSNM. Good agreement was found for the high spectral contrast of gypsum and for the low spectral contrast of water bodies. Furthermore, the day/night consistency of ASTER emissivities was checked, and day/night emissivity differences lower than ±0.013 were observed. However, unexpected emissivity values larger than unity were retrieved by ASTER/TES at 8-9 μm , mainly concentrated over lava flow surfaces. The thermal infrared radiance image data with 90-m spatial resolution was resized to 180 m for the analysis in this paper to avoid misregistration problems due to terrain topography. Emissivity temporal variations were analyzed and attributed, in some cases, to the soil moisture variations. This was particularly noted after periods of high precipitation which occurred in August 2006. The results presented here show the high emissivity accuracy achievable with ASTER data in ideal atmospheric conditions and discuss some problems which should be considered in the future, as the retrieval of overestimated emissivity values.

Comparison of field emissivities with laboratory measurements and ASTER data

Surface emissivity in the thermal infrared (TIR) region is an important parameter for determining the land surface temperature from remote sensing measurements. This work compares the emissivities measured by different field methods (the Box method and the Temperature and Emissivity Separation, TES, algorithm) as well as emissivity data from ASTER scenes and the spectra obtained from the ASTER Spectral Library. The study was performed with a field radiometer having TIR bands with central wavelengths at 11.3 μm, 10.6 μm, 9.1 μm, 8.7 μm and 8.4 μm, similar to the ASTER TIR bands. The measurements were made at two sites in southern New Mexico. The first was in the White Sands National Monument, and the second was an open shrub land in the Jornada Experimental Range, in the northern Chihuahuan Desert, New Mexico, USA. The measurements show that, in general, emissivities derived with the Box method agree within 3% with those derived with the TES method for the spectral bands centered at 10.6 μm and 11.3 μm. However, the emissivities for the shorter wavelength bands are higher when derived with the Box method than those with the TES algorithm (differences range from 2% to 7%). The field emissivities agree within 2% with the laboratory spectrum for the 8-13 μm, 11.3 μm and 10.6 μm bands. However, the field and laboratory measurements in general differ from 3% to 16% for the shorter wavelength bands, i.e., 9.1 μm, 8.6 μm and 8.4 μm. A good agreement between the experimental measurements and the ASTER TIR emissivity data is observed for White Sands, especially over the 9 - 12 μm range (agreement within 4%). The study showed an emissivity increase up to 17% in the 8 to 9 μm range and an increase of 8% in emissivity ratio of average channels (8.4 μm, 8.6 μm, 9.1 μm):(10.6 μm, 11.3 μm) for two gypsum samples with different water content.

Comparison of Modis Emissivity Observation with Laboratory Measurements

The main aim of this work is to analyze the thermal infrared (TIR) emissivity retrievals from the Moderate Resolution Imaging Spectrometer (MODIS) sensor onboard National Aeronautics and Space Administration (NASA)s Terra satellite. The emissivities are compared with the laboratory spectra of the soil samples. Nearly two years of 8-day and monthly composites of TIR surface emissivity data from the day/night LST algorithm were analyzed for temporal variations over White Sands and the Sahara desert. An emissivity increase for band 29 (8.6 mum) was observed at the end of 2005. We believe that the increase is due to soil moisture. The study shows that surface infrared emissivities derived from the MODIS data preserves well the spectral shape of the soils, being the agreement with laboratory measurements about 2 or 3% in average.

Thomas Schmugge awarded 2006 Robert E. Horton medal

Thomas Schmugge was awarded the Robert E. Horton Medal at the AGU Fall Meeting honors ceremony, which was held on 13 December 2006 in San Francisco, Calif. The medal recognizes outstanding contributions to hydrology. For the past 30 years, Thomas Schmugge has been a truly peerless intellectual leader in improving both the theory and the application of microwave and infrared radiative transfer for the remote sensing of the land surface, especially soil moisture, surface temperature, and emissivity. Without exaggeration, he has been a trailblazer and is now unquestionably one of the few world experts on remote sensing in hydrology.