Challenges and uncertainty in plot-scale emissivity and surface temperature estimation using flux tower measurements

Land surface temperature (LST) is a preeminent state variable that controls the energy and water exchange between the Earth’s surface and the atmosphere. At the landscape-scale, LST is derived from thermal infrared radiance measured using space-borne radiometers. At the plot-scale, the flux tower recorded longwave radiation components are inverted to retrieve LST. Since the down-welling longwave component was not measured routinely until recently, usually only the up-welling longwave component is used for the plot-scale LST retrieval. However, we found that ignoring reflected down-welling longwave radiation for plot-scale LST estimations can lead to substantial error. This also has important implications for estimating the correct surface emissivity using flux tower measurements, which is needed for plot-scale LST retrievals. The present study proposes a new method for plot-scale emissivity and LST estimation and addresses in detail the consequences of omitting down-welling longwave radiation as frequently done in the literature. Our analysis uses ten eddy covariance sites with different land cover types and found that the LST values obtained using both up-welling and down-welling longwave radiation components are 0.5 to 1.5 K lower than estimates using only up-welling longwave radiation. Furthermore, the proposed method helps identify inconsistencies between plot-scale radiometric and aerodynamic measurements, likely due to footprint mismatch between measurement approaches. We also found that such inconsistencies can be removed by slight corrections to the up-welling longwave component and subsequent energy balance closure, resulting in realistic estimates of surface emissivity and consistent relationships between energy fluxes and surface-air temperature differences. Landscape-scale day-time LST obtained from satellite data (MODIS TERRA) was strongly correlated with our plot-scale estimates for most of the sites, but higher by several Kelvin at two sites. We also quantified the uncertainty in estimated LST and surface emissivity using the different methods and found that the proposed method does not result in increased uncertainty. The results of this work have significant implications for the combined use of aerodynamic and radiometric measurements to understand the interactions and feedbacks between LST and surface-atmosphere exchange processes.

Challenges and uncertainty in plot-scale emissivity and surface temperature estimation using flux tower measurements

Land surface temperature (LST) is a preeminent state variable that controls the energy and water exchange between the Earth’s surface and the atmosphere. At the landscape-scale, LST is derived from thermal infrared radiance measured using space-borne radiometers. At the plot-scale, the flux tower recorded longwave radiation components are inverted to retrieve LST. Since the down-welling longwave component was not measured routinely until recently, usually only the up-welling longwave component is used for the plot-scale LST retrieval. However, we found that ignoring reflected down-welling longwave radiation for plot-scale LST estimations can lead to substantial error. This also has important implications for estimating the correct surface emissivity using flux tower measurements, which is needed for plot-scale LST retrievals. The present study proposes a new method for plot-scale emissivity and LST estimation and addresses in detail the consequences of omitting down-welling longwave radiation as frequently done in the literature. Our analysis uses ten eddy covariance sites with different land cover types and found that the LST values obtained using both up-welling and down-welling longwave radiation components are 0.5 to 1.5 K lower than estimates using only up-welling longwave radiation. Furthermore, the proposed method helps identify inconsistencies between plot-scale radiometric and aerodynamic measurements, likely due to footprint mismatch between measurement approaches. We also found that such inconsistencies can be removed by slight corrections to the up-welling longwave component and subsequent energy balance closure, resulting in realistic estimates of surface emissivity and consistent relationships between energy fluxes and surface-air temperature differences. Landscape-scale daytime LST obtained from satellite data (MODIS TERRA) was strongly correlated with our plot-scale estimates for most of the sites, but higher by several Kelvin at two sites. We also quantified the uncertainty in estimated LST and surface emissivity using the different methods and found that the proposed method does not result in increased uncertainty. The results of this work have significant implications for the combined use of aerodynamic and radiometric measurements to understand the interactions and feedbacks between LST and surface-atmosphere exchange processes.

Satellite Retrieval of Microwave Land Surface Emissivity under Clear and Cloudy Skies in China Using Observations from AMSR-E and MODIS

Microwave land surface emissivity (MLSE) is an important geophysical parameter to determine the microwave radiative transfer over land and has broad applications in satellite remote sensing of atmospheric parameters (e.g., precipitation, cloud properties), land surface parameters (e.g., soil moisture, vegetation properties), and the parameters of interactions between atmosphere and terrestrial ecosystem (e.g., evapotranspiration rate, gross primary production rate). In this study, MLSE in China under both clear and cloudy sky conditions was retrieved using satellite passive microwave measurements from Aqua Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E), combined with visible/infrared observations from Aqua Moderate Resolution Imaging Spectroradiometer (MODIS), and the European Centre for Medium-Range Weather Forecasts (ECMWF) atmosphere reanalysis dataset of ERA-20C. Attenuations from atmospheric oxygen and water vapor, as well as the emissions and scatterings from cloud particles are taken into account using a microwave radiation transfer model to do atmosphere corrections. All cloud parameters needed are derived from MODIS visible and infrared instantaneous measurements. Ancillary surface skin temperature as well as atmospheric temperature-humidity profiles are collected from ECMWF reanalysis data. Quality control and sensitivity analyses were conducted for the input variables of surface skin temperature, air temperature, and atmospheric humidity. The ground-based validations show acceptable biases of primary input parameters (skin temperature, 2 m air temperature, near surface relative humidity, rain flag) for retrieving using. The subsequent sensitivity tests suggest that 10 K bias of skin temperature or observed brightness temperature may result in a 4% (~0.04) or 7% (0.07) retrieving error in MLSE at 23.5 GHz. A nonlinear sensitivity in the same magnitude is found for air temperature perturbation, while the sensitivity is less than 1% for 300 g/m2 error in cloud water path. Results show that our algorithm can successfully retrieve MLSE over 90% of the satellite detected land surface area in a typical cloudy day (cloud fraction of 64%), which is considerably higher than that of the 29% area by the clear-sky only algorithms. The spatial distribution of MLSE in China is highly dependent on the land surface types and topography. The retrieved MLSE is assessed by compared with other existing clear-sky AMSR-E emissivity products and the vegetation optical depth (VOD) product. Overall, high consistencies are shown for the MLSE retrieved in this study with other AMSR-E emissivity products across China though noticeable discrepancies are observed in Tibetan Plateau and Qinling-Taihang Mountains due to different sources of input skin temperature. In addition, the retrieved MLSE exhibits strong positive correlations in spatial patterns with microwave vegetation optical depth reported in the literature.

Numerical Investigation of the Effects of Window Height and Gas Thickness on Heat Transfer and Gas Flow in Double Pane Windows

Double pane window is an effective way to reduce the heat loss from windows in buildings. There are many studies on the thermal performance of these window applications for different parameters such as optimum gap width, suitable filling fluid and different applications such as film coatings on panes to obtain different surface emissivity values or placing venetian blinds inside the gap, etc. These investigations are mostly based on the laminar flow assumption inside the gas gap between the two panes for the same window height. In this research, effect of the window height and gap width on the gas flow in the gap and heat transfer over double pane for three cities of Turkey representing different climates were numerically investigated with turbulent flow and ideal gas assumptions inside the gap for air and argon. In the calculations, natural convection for pane surface facing indoors and forced convection for pane surface facing outdoors was assumed as boundary condition. The numerical results shown that also the window height such as gap width has an effect on the heat transfer and gas flow of the double pane window. Thereby, the window height should be taken into consideration for determining the optimum gap width in the double pane window applications.

A novel method using infrared thermography for hot fluid leakage detection on surfaces with uneven emissivities

Uneven surface emissivity will cause illusory temperature variation in infrared surface temperature mapping. For this reason, most of the detailed reviews on the use of infrared thermography (IRT) for leakage detection have mainly focused on surfaces with homogeneous emissivity or the recognition of negative temperature gradients, while reports on sensing hot fluid leakage for uneven surface emissivity are very rare. In this study, a hypothesis is put forward and a new leakage detection method is proposed that uses a transient heating-cooling-heating process in association with a subtraction method of infrared images to eliminate the disturbance of inhomogeneous valve surface emissivities. A theoretical analysis is established that is experimentally tested as a case study. The results shows that the hypothesis is clear and the effect of the uneven emissivity is suppressed for the recognition of positive temperature gradients (hot fluid leakage) on a metal valve sample. The current work provides new insights on the modification of the surface emissivity under certain conditions, which has been a major limitation of passive IRT in the past.