Retrieval of atmospheric and land surface parameters from satellite-based thermal infrared hyperspectral data using a neural network technique

Chen, Mengshuo; Li, Ni; Jiang, Xuefeng; Li, Zhao-Liang; Wu, Hua

doi:10.1080/01431161.2012.716536

Cited by 45 publications

(23 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…LSE has been estimated for several decades using many previously developed methods (Becker and Li 1990;Li and Becker 1993;Valor and Caselles 1996;Wan and Li 1997;Snyder et al 1998;Sobrino and Raissouni 2000;Li, Petitcolin, and Zhang 2000;Jiang, Li, and Nerry 2006;Bi 2009, 2011;Qian et al 2013;Wang et al 2013;Gao et al 2014). comprehensively reviewed these different methods and grouped them into three categories: (1) (semi-)empirical or theoretical methods, (2) multi-channel temperature emissivity separation (TES) methods, and (3) physically based methods (PBMs).…”

Section: Introductionmentioning

confidence: 99%

An improved NDVI-based threshold method for estimating land surface emissivity using MODIS satellite data

Tang

Shao

et al. 2015

International Journal of Remote Sensing

View full text Add to dashboard Cite

This work estimated the land surface emissivities (LSEs) for MODIS thermal infrared channels 29 (8.4-8.7 μm), 31 (10.78-11.28 μm), and 32 (11.77-12.27 μm) using an improved normalized difference vegetation index (NDVI)-based threshold method. The channel LSEs are expressed as functions of atmospherically corrected reflectance from the MODIS visible and near-infrared channels with wavelengths ranging from 0.4 to 2.2 μm for bare soil. To retain the angular information, the vegetation LSEs were explicitly expressed in the NDVI function. The results exhibited a root mean square error (RMSE) among the estimated LSEs using the improved method, and those calculated using spectral data from Johns Hopkins University (JHU) are below 0.01 for channels 31 and 32. The MODIS land surface temperature/emissivity (LST/E) products, MOD11_L2 with LSE derived via the classification-based method with 1 km resolution and MOD11C1 with LSE retrieved via the day/night LST retrieval method at 0.05°resolution, were used to validate the proposed method. The resultant variances and entropies for the LSEs estimated using the proposed method were larger than those extracted from MOD11_L2, which indicates that the proposed method better described the spectral variation for different land covers. In addition, comparing the estimated LSEs to those from MOD11C1 yielded RMSEs of approximately 0.02 for the three channels; however, more than 70% of pixels exhibited LSE differences within 0.01 for channels 31 and 32, which indicates that the proposed method feasibly depicts LSE variation for different land covers.

show abstract

Section: Introductionmentioning

confidence: 99%

An improved NDVI-based threshold method for estimating land surface emissivity using MODIS satellite data

Tang

Shao

et al. 2015

International Journal of Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Inspired by several previous studies [36][37][38][39], a modified split-window covariance and variance ratio (MSWCVR) method was developed to retrieve CWV from the TIRS data. With a vital assumption that the atmosphere is unchanged over the neighboring pixels, the MSWCVR method relates the atmospheric CWV to the ratio of the upward transmittances in two thermal infrared bands, whereas the transmittance ratio can be calculated based on the TOA brightness temperatures of the two bands.…”

Section: Determination Of Atmospheric Cwvmentioning

confidence: 99%

A Practical Split-Window Algorithm for Estimating Land Surface Temperature from Landsat 8 Data

et al. 2015

View full text Add to dashboard Cite

This paper developed a practical split-window (SW) algorithm to estimate land surface temperature (LST) from Thermal Infrared Sensor (TIRS) aboard Landsat 8. The coefficients of the SW algorithm were determined based on atmospheric water vapor sub-ranges, which were obtained through a modified split-window covariance-variance ratio method. The channel emissivities were acquired from newly released global land cover products at 30 m and from a fraction of the vegetation cover calculated from visible and near-infrared images aboard Landsat 8. Simulation results showed that the new algorithm can obtain LST with an accuracy of better than 1.0 K. The model consistency to the noise of the brightness temperature, emissivity and water vapor was conducted, which indicated the robustness of the new algorithm in LST retrieval. Furthermore, based on comparisons, the new algorithm performed better than the existing algorithms in retrieving LST from TIRS data. Finally, the SW algorithm was proven to be reliable through application in different regions. To further confirm the credibility of the SW algorithm, the LST will be validated in the future.

show abstract

“…Recently, Suess et al () proposed import vector machines that have been found very effective for quantitative analysis of hyperspectral data. An effective neural network method was proposed for retrieval of atmospheric and land surface parameters from satellite‐based thermal infrared hyperspectral data (Wang et al ).…”

Section: Discussionmentioning

confidence: 99%

Hyperspectral Remote Sensing Classifications: A Perspective Survey

Chutia

Bhattacharyya

Sarma

et al. 2015

Transactions in GIS

View full text Add to dashboard Cite

Classification of hyperspectral remote sensing data is more challenging than multispectral remote sensing data because of the enormous amount of information available in the many spectral bands. During the last few decades, significant efforts have been made to investigate the effectiveness of the traditional multispectral classification approaches on hyperspectral data. Formerly extensively established conventional classification methods have been dominated by the advanced classification approaches and many pre-processing techniques have been developed and incorporated in hyperspectral classification. A perspective survey of hyperspectral remote sensing classification approaches is presented here. It comprehensively highlights the taxonomy of major classification approaches reported during the last two decades and describes an experimental evaluation of a few major classification algorithms. Recent advancements in the development of classification approaches are also emphasized with a set of guidelines for achieving better classification performances.

show abstract

Retrieval of atmospheric and land surface parameters from satellite-based thermal infrared hyperspectral data using a neural network technique

Cited by 45 publications

References 22 publications

An improved NDVI-based threshold method for estimating land surface emissivity using MODIS satellite data

An improved NDVI-based threshold method for estimating land surface emissivity using MODIS satellite data

A Practical Split-Window Algorithm for Estimating Land Surface Temperature from Landsat 8 Data

Hyperspectral Remote Sensing Classifications: A Perspective Survey

Contact Info

Product

Resources

About