Joan M. Galve

Validation of Landsat-7/ETM+ Thermal-Band Calibration and Atmospheric Correction With Ground-Based Measurements

Ground-based measurements of land-surface temperature (LST) performed in a homogeneous site of rice crops close to Valencia, Spain, were used for the validation of the calibration and the atmospheric correction of the Landsat-7 Enhanced Thematic Mapper Plus (ETM+) thermal band. Atmospheric radiosondes were launched at the test site around the satellite overpasses. Field-emissivity measurements of the near-full-vegetated rice crops were also performed. Seven concurrences of Landsat-7 and ground data were obtained in July and August 2004-2007. The ground measurements were used with the MODTRAN-4 radiative transfer model to simulate at-sensor radiances and brightness temperatures, which were compared with the calibrated ETM+ observations over the test site. For the cases analyzed here, the differences between the simulated and ETM+ brightness temperatures show an average bias of 0.6 K and a rootmean-square difference (rmsd) of ±0.8 K. The ground-based measurements were also used for the validation of LSTs derived from ETM+ at-sensor radiances with atmospheric correction calculated from the following: 1) the local-radiosonde profiles and 2) the operational atmospheric-correction tool available at http://atmcorr.gsfc.nasa.gov. For the first case, the differences between the ground and satellite LSTs ranged from -0.6 to 1.4 K, with a mean bias of 0.7 K and an rmsd = ±1.0 K. For the second case, the differences ranged between -1.8 and 1.3 K, with a zero average bias and an rmsd = ±1.1 K. Although the validation cases are few and limited to one land cover at morning and summer, results show the good LST accuracy that can be achieved with ETM+ thermal data.

An Atmospheric Radiosounding Database for Generating Land Surface Temperature Algorithms

A database of global, cloud-free, and atmospheric radiosounding profiles was compiled with the aim of simulating radiometric measurements from satellite-borne sensors in the thermal infrared. The objective of the simulated data is to generate split-window (SW) and dual-angle (DA) algorithms for the retrieval of land surface temperature (LST) from Terra/Moderate Resolution Imaging Spectroradiometer (MODIS) and Envisat/Advanced Along Track Scanning Radiometer (AATSR) data. The database contains 382 radiosounding profiles acquired over land, with nearly uniform distribution of precipitable water between 0.02 and 5.5 cm. Radiative transfer calculations were performed with the MODTRAN 4 code for six viewing angles between 0deg and 60deg. The resulting radiance spectra were convoluted with the response filter functions of MODIS bands 31 and 32 and AATSR channels at 11 and 12 mum. By using the simulation database, the SW algorithms adapted for MODIS and AATSR data and the DA algorithms for AATSR data were developed. Both types of algorithms are quadratic in the brightness temperature difference and depend explicitly on the land surface emissivity. The SW and DA algorithms were validated with actual ground measurements of LST collected concurrently to MODIS and AATSR observations in a site located close to the city of Valencia, Spain, in a large, flat, and thermally homogeneous area of rice crops. The results obtained have no bias and a standard deviation around plusmn0.5 K for the SW algorithms at nadir for both sensors. The SW algorithm used in the forward view results in a bias of 0.6 K and a standard deviation of plusmn0.8 K. The worst results are obtained in the other algorithms with a bias close to -1.0 K and a standard deviation close to plusmn1.1 K in the case of the DA algorithms.

Soil Moisture Effect on Thermal Infrared (8–13-μm) Emissivity

Thermal infrared (TIR) emissivities of soils with different textures were measured for several soil moisture (SM) contents under controlled conditions using the Box method and a high-precision multichannel TIR radiometer. The results showed a common increase of emissivity with SM at water contents lower than the field capacity. However, this dependence is negligible for higher water contents. The highest emissivity variations were observed in sandy soils, particularly in the 8-9-μm range due to water adhering to soil grains and decreasing the reflectance in the 8-9-μm quartz doublet region. Thus, in order to model the emissivity dependence on soil water content, different approaches were studied according to the a priori soil information. Soil-specific relationships were provided for each soil texture and different spectral bands between 8 and 13 μm, with determination coefficients up to 0.99, and standard estimation errors in emissivity lower than ± 0.014. When considering a general relationship for all soil types, standard estimation errors up to ±0.03 were obtained. However, if other soil properties (i.e., organic matter, quartz, and carbonate contents) were considered, along with soil water content, the general relationship predicted TIR emissivities with a standard estimation error of less than ±0.008. Furthermore, the study showed the possibility of retrieving SM from TIR emissivities with a standard estimation error of about ±0.08 m3 . m-3.

Land Surface Temperature From the Advanced Along-Track Scanning Radiometer: Validation Over Inland Waters and Vegetated Surfaces

The land surface temperature (LST) product of the Advanced Along-Track Scanning Radiometer (AATSR) was validated with ground measurements at the following two thermally homogeneous sites: Lake Tahoe, CA/NV, USA, and a large rice field close to Valencia, Spain. The AATSR LST product is based on the split-window technique using the 11- and 12- mum channels. The algorithm coefficients are provided for 13 different land-cover classes plus one lake class (index i). Coefficients are weighted by the vegetation-cover fraction (f). In the operational implementation of the algorithm, i and f are assigned from a global classification and monthly fractional vegetation-cover maps with spatial resolutions of 0.5deg times 0.5deg. Since the validation sites are smaller than this, they are misclassified in the LST product and treated incorrectly despite the fact that the higher resolution AATSR data easily resolve the sites. Due to this problem, the coefficients for the correct cover types were manually applied to the AATSR standard brightness temperature at sensor product to obtain the LST for the sites assuming they had been correctly classified. The comparison between the ground-measured and the AATSR-derived LSTs showed an excellent agreement for both sites, with nearly zero average biases and standard deviations les 0.5degC. In order to produce accurate and precise estimates of LST, it is necessary that the land-cover classification is revised and provided at the same resolution as the AATSR data, i.e., 1 km rather than the 0.5deg resolution auxiliary data currently used in the LST product.

Evaluation of Different Methods to Retrieve the Hemispherical Downwelling Irradiance in the Thermal Infrared Region for Field Measurements

The thermal infrared hemispherical downwelling irradiance (HDI) emitted by the atmosphere and surrounding elements contributes through reflection to the signal measured over an observed surface by remote sensing. This irradiance must be estimated in order to obtain accurate values of land-surface temperature (LST). There are some fast methods to measure the HDI with a single measurement pointing to the sky at a specified viewing direction, but these methods require completely cloud-free or cloudy skies, and they do not account for the radiative contribution of surrounding elements. Another method is the use of a diffuse reflectance panel (usually, a rough gold-coated surface) with near-Lambertian behavior. This method considers the radiative contribution of surrounding elements and can be used under any sky condition. A third possibility is the use of atmospheric profiles and a radiative transfer code (RTC) in order to simulate the atmospheric signal and to calculate the HDI by integration. This study compares the HDI estimations with these approaches, using measurements made on four different days with a completely clear sky and two days with a partially cloudy sky. The measurements were made with a four-channel CIMEL Electronique radiometer working in the 8-14-μm spectral range. The HDI was also estimated by means of National Centers for Environmental Prediction atmospheric profiles introduced in the MODTRAN RTC. Additionally, the measurements were made at two different places with very different environments to quantify the effect of the contributing surroundings. Results showed that, for a clear-sky day with a minimal contribution of the surroundings, all methods differed from each other between 5% and 11%, depending on the spectral range, and any of them could be used to estimate HDI in these conditions. However, in the case of making surface measurements in an area with significant surrounding elements (buildings, trees, etc.), HDI values retrieved from the panel present an increase of +3 W·m-2·μm-1 compared with the other methods; this increase, if ignored, implies to make an error in LST ranging from +0.5 °C to +1.5 °C, depending on the spectral range and on surface emissivity and temperature. Comparison under heterogeneous skies with changing cloud coverage showed also large differences between the use of panel and the other methods, reaching a maximum difference of +4.6 W·m-2·μm-1, which implies to make an error on LST of +2.2 °C. In these cases, the use of the diffuse reflectance panel is proposed, since it is the unique way to capture the contribution of the surroundings and also to adequately measure HDI for sky changing conditions.

Angular dependence of the emissivity of bare soils in the thermal infrared

Emissivity is one of the main factors to take into account when studying processes that take place in the Earth surface by using radiance measurements in the thermal infrared, such as surface energy balance, land surface temperature (LST) retrieval, classification of different types of surface, etc. For this reason it is necessary to study the factors that can influence the emissivity. The present work evaluates one of these factors: the variation of the emissivity with the zenithal observation angle over bare soils, specifically the variation of the relative emissivity calculated from measurements of radiances, almost simultaneous, at nadir (0o) and at a certain angle (¿). The measurements of radiance were taken with the aid of a straightforward goniometric system that allows the measurement from nadir observation to 70o (at 10o increments) for a fixed azimuthal angle. The results show a significant decrease of emissivity with observation angle, which is especially accentuated in the case of sandy soils with high quartz content.

Estimation of atmospheric water vapour content from direct measurements of radiance in the thermal infrared region

Atmospheric water vapour content is a required parameter in thermal infrared (TIR) to carry out processes such as atmospheric correction or retrieving atmospheric factors (downwelling or upwelling irradiance, transmittance of the atmosphere and so on). This study proposes an alternative method to the ones already in use to measure water vapour content from direct measurements of downwelling atmospheric radiance in the TIR range. It was possible to estimate a linear relationship between atmospheric water vapour and downwelling atmospheric radiance using a simulated study, based on data from a radiosounding database. A subsequent validation concludes that it is possible to obtain water vapour content with an uncertainty of 0.5 cm using in situ measurements of downwelling atmospheric radiance in the TIR range of 11.5–12.5 μm.

An enhanced vegetation cover method for automatic generation of land surface emissivity maps

The remote sensing measurement of the land surface temperature from satellites provides an overview of this magnitude on a continuous and regular basis. The study of its evolution in time and space is a critical factor in many scientific fields such as weather forecasting, detection of forest fires, climate change, and so on. The main problem of making this measurement from satellite data is the need to correct the effects of the atmosphere and the surface emissivity. The aim was to define an enhanced vegetation cover method to calculate and generate, automatically, maps of land surface emissivity from images of the AATSR (Advanced Along Track Scanning Radiometer) onboard the Envisat satellite. For the production of these emissivity maps, the geometric model purposed is based on [6]. Its validation was made by comparing the obtained results and the values measured in previous field campaigns [2] carried out in the area of rice fields of Valencia, Spain.

Single band atmospheric correction tool for thermal infrared data: application to Landsat 7 ETM+

Atmospheric correction of Thermal Infrared (TIR) remote sensing data is a key process in order to obtain accurate land surface temperatures (LST). Single band atmospheric correction methods are used for sensors provided with a single TIR band. Which employs a radiative transfer model using atmospheric profiles over the study area as inputs to estimate the atmospheric transmittances and emitted radiances. Currently, TIR data from Landsat 5-TM, Landsat 7-ETM+ and Landsat 8-TIRS can be atmospherically corrected using the on-line Atmospheric Correction Parameter Calculator (ACPC, http://atmcorr.gsfc.nasa.gov). For specific geographical coordinates and observation time, the ACPC provides the atmospheric transmittance, and both upwelling and downwelling radiances, which are calculated from MODTRAN4 radiative transfer simulations with NCEP atmospheric profiles as inputs. Since the ACPC provides the atmospheric parameters for a single location, it does not account for their eventual variability within the full Landsat scene. The new Single Band Atmospheric Correction (SBAC) tool provides the geolocated atmospheric parameters for every pixel taking into account their altitude. SBAC defines a three-dimensional grid with 1°×1° latitude/longitude spatial resolution, corresponding to the location of NCEP profiles, and 13 altitudes from sea level to 5000 meters. These profiles are entered in MODTRAN5 to calculate the atmospheric parameters corresponding to a given pixel are obtained by weighted spatial interpolation in the horizontal dimensions and linear interpolation in the vertical dimension. In order to compare both SBAC and ACPC tools, we have compared with ground measurements the Landsat-7/ETM+ LST obtained using both tools over the Valencia ground validation site.

Comparison of Split-Window and Single-Channel Methods for Land Surface Temperature Retrieval from MODIS and AATSR Data

In this study, two different methods for retrieving the Land Surface Temperature (LST) from Terra/Moderate Resolution Imaging Spectroradiometer (MODIS) and Envisat/Advanced Along Track Scanning Radiometer (AATSR) data are compared against a database of ground measured LSTs. These are the split-window (SW) and the single-channel (SC) methods. The SW method expresses LST as a combination of the brightness temperatures in the 11 iquestm and 12 iquestm channels with coefficients that can have local or global validity, depending on the way they are obtained. SC methods are based on the atmospheric radiative transfer equation. To solve this equation, convenient atmospheric temperature and water vapor profiles are required as inputs of a radiative transfer model. In this work we used three different sources of atmospheric profiles: local radiosoundings (LR) reanalysis model output (RM), and satellite-based profiles (SB). Results show that the SW method produces more accurate LST retrievals (plusmn0.5 K) than the SC method with the three profile sources used, of which the RM source yielded the best results (plusmn1.0 K).