Advances in Environmental Remote Sensing Sensors, Algorithms, and Applications

Synopsis of the Book

This book is divided into four sections. Section I deals with various sensors, systems, or sensing using different regions of wavelengths. Section II exemplifies recent advances in algorithms and techniques, specifically in image preprocessing and thematic information extraction. Section III focuses on remote sensing of vegetation and related features of the Earth’s surface. Finally, Section IV examines developments in the remote sensing of air, water, and other terrestrial features.

The chapters in Section I provide a comprehensive overview of some important sensors and remote sensing systems, with the exception of Chapter 5. By reviewing key concepts and methods and illustrating practical uses of particular sensors/sensing systems,these chapters provide insights into the most recent developments and trends in remote sensing and further identify the major existing problems of these trends. These remote sensing systems utilize visible, reflected infrared, thermal infrared, and microwave spectra,and include both passive and active sensors. In Chapter 1, Song and his colleagues evaluate one of the longest remote sensing programs in the world, that is, the U.S. Landsat program, and discuss its applications in vegetation studies. With a mission of long-term monitoring of vegetation and terrestrial features, Landsat has built up a glorious history.

The remote sensing literature is filled with a large number of articles in vegetation classification

and change detection. However, remote sensing of vegetation remains a great challenge, especially the sensing of biophysical parameters such as leaf area index (LAI),biomass, and forest successional stages (Song, Gray, and Gao 2010). A remarkable strength of the Landsat program is its time-series data, especially when considering the additionof the upcoming Landsat Data Continuity Mission (LDCM); however, these data are not a panacea for vegetation studies. Song, Gray, and Gao (2010) suggest that the synergistic use of data from other remote sensors may provide complimentary vegetation information to Landsat data, such as high spatial resolution (<10 m) satellite images that provide textural information, radar sensors that provide information on the dielectric properties of the surface and are capable of penetrating clouds, light detection and ranging (LiDAR, which provides

geometric information), and coarse spatial but high temporal resolution sensors (e.g., Moderate Resolution Imaging Spectroradiometer [MODIS]). Chapter 1 provides an excellent example for the integrated use of Landsat and MODIS data by introducing the spatial and temporal adaptive reflectance fusion model (Chapter 1, Section 1.3.5; Gao et al. 2006).

In Chapter 2, Shao and his colleagues provide a comprehensive review of selected data products, algorithms, and applications of MODIS. MODIS has its roots in earlier sensors such as the Advanced Very High Resolution Radiometer (AVHRR) and coastal zone color scanner (CZCS), but provides substantial improvements over these earlier sensing systems (Lillesand, Kiefer, and Chipman 2008). MODIS provides a wide range of data products applicable to land, ocean, and atmosphere. Chapter 2 focuses on the examination of land products and applications, in particular, application studies at the global and regional levels. For each data product, the contributors document most recent advances, but also point out the product’s limitations in data quality and validation.

Lidar has been increasingly used in many geospatial applications due to its high data resolution, low consumption of time and cost, compared to many traditional remote sensing technologies. Unlike other remotely sensed data, LiDAR data focus solely on geometry rather than on radiometry. Many researchers have used LiDAR in conjunction with optical remote sensing and geographic information system (GIS) data in urban, environment, and resource studies (Weng 2009). Chapter 3 offers a detailed introduction of the basic concept of LiDAR, and types of sensors and platforms.


Based on the works of the author this chapter further provides a review of LiDAR remote sensing applications in estimating forest biophysical parameters and surface and canopy fuels, and for characterizing wildlife habitats.Synthetic aperture radar (SAR) has been a key sensing system for various environmental applications, and the Earth and planetary exploration. 

In Chapter 4, Franceschetti and Tatoian introduce to the reader two new concepts of SAR imaging: (1) impulse SAR and (2) polychromatic SAR. The theoretical foundations of the two systems are presented with some preliminary experimental data for validating the theory. The authors further discuss the distinct advantages of these systems over conventional microwave imaging sensors and their potential applications, and speculate on future research directions. Hyperspectral remote sensing, as a cutting-edge technology, has been widely applied in vegetation and ecological studies.

Chapter 5 provides an overview of spectral characteristics for a set of plant biophysical and biochemical parameters. A wide range of techniques are reviewed, including such spectral analysis techniques as spectral derivative analysis, spectral matching, spectral index analysis, spectral absorption features and spectral position variables, hyperspectral transformation, spectral unmixing analysis, and hyperspectral classifications. Further, two general analytical approaches are discussed: (1) empirical/ statistical methods and (2) physically based modeling. The chapter concludes with the authors’ perspectives on the future directions of hyperspectral remote sensing of vegetation biophysical parameters. Thermal infrared (TIR) remote sensing techniques have been applied in urban climate and environmental studies. 

Chapter 6 examines the current practices, problems, and prospects of this particular field of study, especially the applications of remotely sensed TIR data in urban studies. It is suggested that the majority of previous researches have focused on land-surface temperature (LST) patterns and their relationships with urban-surface biophysical characteristics, especially with vegetation indices and land-use/land-cover types. Less attention has been paid to the derivation of urban heat island (UHI) parameters from LST data and to the use of remote sensing techniques to estimate surface energy fluxes.


Major recent advances, future research directions, and the impacts of planned TIR sensors with LDCM and HyspIRI missions are outlined in the chapter.Section II presents new developments in algorithms and techniques, specifically in image preprocessing, thematic information extraction, and digital change detection. 

Chapter 7 conducts a concise review of atmospheric correction algorithms for the optical remote sensing of land. This review focuses on physical models of atmospheric correction that describe the radiative transfer in the Earth’s atmosphere, instead of empirical methods.The author presents sequentially the correction algorithms for hyperspectral, thermal, and multispectral sensors, then discusses the combined method for performing topographic and atmospheric corrections, and ends with examples of correcting non-standard atmospheric conditions, including haze, cirrus, and cloud shadow. The chapter concludes with the author’s perspective on major challenges and future research needs in atmospheric and topographic correction. 

In addition, the chapter includes a brief survey and a comparison of capacity among commercially available atmospheric correction software/modules, which will be very useful for students.Geometric correction is more important now than ever due mainly to the growing need for off-nadir and high-resolution imaging, fully digital processing and interpretation of remote sensing images, and image fusion and remote sensing–GIS data integration in practical applications (Toutin 2010). Three-dimensional (3D) geometric processing and correction of Earth observation (EO) satellite data is a key issue in multisource, multiformat data integration, management, and analysis for many EO and geomatic applications (Toutin 2010). 

Chapter 8 first reviews the source of geometric distortions (with relation to platform,sensor, other measuring instruments, Earth, and atmosphere), and then compares different mathematical models for correcting geometric distortions (e.g., 2D/3D polynomial, 3D rational functions, and physical and deterministic models). Subsequently, the methods and algorithms in each processing step of the geometric correction are examined in detail,supplemented with plentiful literature. 

This type of examination allows the tracking of error propagation from the input data to the final output product.Image classification is a fundamental protocol in digital image processing and provides crucial information for subsequent environmental and socioeconomic applications.Generating a satisfactory classification image from remote sensing data is not a straightforward task. Many factors contribute to this difficulty, including the characteristics of a study area, availability of suitable remote sensing data, ancillary and ground reference data, proper use of variables and classification algorithms, and the analyst’s experience (Lu and Weng 2007). 

Chapter 9 provides a brief overview of the major steps in image classification, and examines the techniques for improving classification performance, including the use of spatial information, multitemporal and ancillary data, and image fusion. A case study is further presented that explores the role of vegetation indices and textural images in improving vegetation classification performance in a moist tropical region of the Brazilian Amazon with Landsat Thematic Mapper (TM) imagery.

Object-based image analysis (OBIA; or GEOBIA for geospatial OBIA) is becoming a new paradigm among the mapping sciences (Blaschke 2010). With the improvement of OBIA software capacity and the increased availability of high spatial resolution satellite images and LiDAR data, vegetation-mapping capabilities are expected to grow rapidly in the nearfuture in terms of both the accuracy and the amount of biophysical vegetation parameters that can be retrieved (Blaschke, Johansen, and Tiede 2010). 

Chapter 10 reviews the development of OBIA and the current status of its application in vegetation mapping. Two case studies are provided to illustrate this mapping capacity. The first case uses LiDAR data to map riparian zone extent and to estimate plant project cover (PPC) within the riparian zone in central Queensland, Australia. Whereas PPC was calculated at the pixel level, OBIA was used for mapping the riparian zone extent and validating the PPC results. The second case study aims at extracting individual tree crowns from a digital surface model (DSM) by using OBIA and grid computing techniques in the federal state of Upper Austria, Austria. Finally, the contributors share their insights on the existing problems and development trends of OBIA with respect to automation, the concept of scale, transferability of rules,and the impacts of improved remote sensing capacities.

Digital change detection requires the careful design of each step, including the statement of research problems and objectives, data collection, preprocessing, selection of suitable detection algorithms, and evaluation of the results (Lu et al. 2010). Errors or uncertainties may emerge from any of these steps, but it is important to understand the relationship among these steps and to identify the weakest link in the image-processing chain (Lu et al. 2010). 

In Chapter 11, Lu and his colleagues update earlier research (Lu et al. 2004) by re-examining the essential steps in change detection and by providing a case study for detecting urban land-use/land-cover in a complex urban–rural frontier in Mato Grosso state, Brazil, based on the comparison of extracted impervious surface data from multitemporal Landsat TM images. They conclude that the selection of a change detection procedure, whether a per-pixel, a subpixel, or an object-oriented method, must conform to the research objectives, remote sensing data used, and geographical size of the study area.The remaining sections of the book focus on various environmental applications of remote sensing technology. 

Section III centers on the remote sensing of vegetation, but each chapter has a very different approach or perspective. Chapter 12 reviews many of the advancements made in the remote sensing of ecosystem structure, processes, and function, and also notes that there exist important trade-offs and compromises in characterizing ecosystems from space related to spatial, spectral, and temporal resolutions of the imaging sensors. Huete and Glenn (2010) suggest that an enormous mismatch exists between leaf-level and species-level ecological variables and satellite spatial resolutions, and this mismatch makes it difficult to validate satellite-derived products. 

They further assert that high temporal resolution hyperspectral remote sensing satellite measurements provide powerful monitoring tools for the characterization of landscape phenology and ecosystem processes, especially when these remote sensing measurements are used in conjunction with calibrated, time-series-based in situ data sets from surface sensor networks.In the western United States, wildfire is a major threat to both humans and the natural environment. Dr. Steve Yool and his colleagues at the University of Arizona have been taking great efforts to study the dynamic relationships among fire, climate, and people from an interdisciplinary perspective, which has been termed “pyrogeography” (Yool 2009). 

In Chapter 13, Yool introduces a remote sensing method to estimate and to map a fuel moisture stress index by standardizing normalized difference vegetation index (NDVI) with the Z transform. This index can be employed as a spatial and temporal fine-scale metric to determine fire season (Yool 2010). Based on a case study conducted in southeastern Arizona, the author demonstrate that the onset and length of the fire season depend on elevation and other microclimatic factors. Fire-season summary maps derived from the fuel moisture stress index may potentially provide lead time to plan for future fire seasons (Yool 2010).Knowledge of forest disturbance and regrowth has obvious scientific significance in the context of global environmental change. Forest change analysis by using time-series analysis of Landsat images is a logical approach, given the long history of Landsat data

records (see Chapter 1 for details). 

Chapter 14 introduces an approach for reconstructing forest disturbance history using Landsat data records. Major steps include the development of Landsat time-series stacks (Huang et al. 2009), and performing change analysis using vegetation-change tracker algorithm (Huang et al. 2010). This approach has been used to produce disturbance products for many areas in the United States (Huang 2010). The author thus further presents two examples of application of this approach to the states of Mississippi and Alabama and the seven national forests in the eastern United States.


The application of this approach for an area outside the United States is possible if the area has a long-term satellite data record of quality and temporally frequent acquisitions, and an inventory of Landsat holdings at international ground-receiving stations (Huang 2010).

Satellite-based modeling of the gross primary production (GPP) of terrestrial ecosystems requires high-quality satellite data, extensive field measurements, and effective radiative transfer models. Current satellite-based GPP models are largely founded on the concept of light-use efficiency (Xiao et al. 2010). Such production efficiency models (PEMs) may be grouped into two categories based on how they calculate the absorption of light for photosynthesis: (1) those models using the fraction of photosynthetically active radiation absorbed by vegetation canopy, and (2) those using the fraction of photosynthetically active radiation absorbed by chlorophyll (Xiao et al. 2010). 

Chapter 15 provides a review of satellite-based PEMs and highlights the major differences between these two approaches.The authors conclude that further research efforts are needed in the validation of satellite based production efficiency models (PEMs) and the error reduction of GPP estimates from net ecosystem exchange (NEE) data using a consistent method.In Chapter 16, Thenkabail and colleagues discuss the maps and statistics of global croplands and the associated water use determined by remote sensing and non remote-sensing approaches. Sources of uncertainty in the areas and limitations of existing cropland maps are further examined. Thenkabail et al. (2010) conclude that among four major cropland area maps and statistics at the global level, one study employed a mainly multisensor remote sensing approach, whereas the others used a combination of national statistics and geospatial techniques. 

However, the uncertainties in these major maps and statistics, as well as the geographic locations of croplands, are quite high. They suggest that it is necessary to utilize higher spatial and temporal resolution satellite images to generate global cropland maps with greater geographic precision, crop types, and cropping intensities. Section V presents examples of applications of remote sensing technology for studies of air, water, and land. This section starts with atmospheric remote sensing, which has great significance in the estimation of aerosol and microphysical properties of the atmosphere in order to understand aerosol climatic issues at scales ranging from local and regional to

global. Aerosol monitoring at the local scale is more challenging due to relatively weak atmospheric signals, coarse spatial resolution images, and the spectral confusion between urban bright surfaces and aerosols. 

Chapter 17 reviews MODIS algorithms for aerosol retrieval at both global and local scales, and illustrates them with a research involving the retrieval of aerosol optical thickness (AOT) over Hong Kong and the Pearl River Delta region, China, by using 500-m MODIS data. The feasibility of using 500-m AOT for mapping urban anthropogenic emissions, monitoring changes in regional aerosols, and pinpointing biomass-burning locations is also demonstrated. Wong and Nichol (2010) suggest that due to the high temporal resolution of MODIS imagery, aerosol retrieval can be accomplished on a routine basis for the purpose of air quality monitoring over megacities.The quality of inland, estuarine, and coastal waters is of high ecological and economical importance (Gitelson et al. 2010). 

Chapter 18 demonstrates the development, evaluation, and validation of algorithms for the remote estimation of chlorophyll-a (Chl-a) concentration in turbid, productive, inland, estuarine, and coastal waters, a pigment universally found in all phytoplankton species and routinely used as a substitute for biomass in all types of aquatic environments. The rationale behind the bio-optical algorithms is presented and the suitability of the developed algorithms for accurate estimation of Chl-a concentration is examined. Gitelson et al. (2010) assert that their algorithms, which are developed by a semi-analytical method and calibrated in a restricted geographic area, can be applied to diverse aquatic ecosystems without the need for further parameterization. 

Chapter 19 is concerned with the interaction between the Earth’s land surface and the atmosphere. Here, Petropoulos and Carlson provide a concise review of the development of remote sensing-based methods currently used in the estimation of surface energy fluxes, that is, the one-layer model, two-layer model, and the “triangle” method (Gillies and Carlson 1995; Gillies et al. 1997), by examining the main characteristics and by comparing their strengths and limitations. Next, remote sensing methods for estimation of soil-water content are assessed, which use visible, TIR, and microwave data, or their combinations. 

The remaining half of this chapter provides a detailed account of the triangle method, its theoretical background, implementation, and validation; and the soil–vegetation–atmosphere transfer (SVAT) model, which is essential for the implementation of the protocol.Urban environmental problems have become unprecedentedly significant in the twenty first century. The National Research Council Decadal Survey suggests that urban environment should be defined as a “new science” to be focused on the U.S. satellite missions of the near future (National Research Council 2007). As such, remote sensing of urban and suburban areas has recently become a new scientific frontier (Weng and Quattrochi 2006). 

Chapter 20 reviews remote sensing approaches to measure the biophysical features of the urban environment, and examines the most important concepts and recent research progresses.

This chapter ends with the author’s prospects on future developments and emerging trends in urban remote sensing, particularly, in the aspect of algorithms. The U.S. Geological Survey (USGS) National Land-Cover Database (NLCD) has been developed over the past two decades. NLCD products provide timely, accurate, and spatially explicit national land cover at 30-m resolution, and have proven effective for addressing issues such as ecosystem health, biodiversity, climate change, and land management policy. 

Chapter 21 summarizes major scientific and technical issues in the development of NLCD 1992, NLCD 2001, and NLCD 2006 products. Experiences and lessons learned from the development of NLCD in terms of project design, technical approaches, and project

implementation are documented. Further, future improvements are discussed for the development of next-generation NLCD products, that is, the NLCD 2011.

Content of the Book
Section I Sensors, Systems, and Platforms
1. Remote Sensing of Vegetation with Landsat Imagery.................................................... 3
Conghe Song, Joshua M. Gray, and Feng Gao
2. Review of Selected Moderate-Resolution Imaging Spectroradiometer
Algorithms, Data Products, and Applications................................................................. 31
Yang Shao, Gregory N. Taff, and Ross S. Lunetta



3. Lidar Remote Sensing........................................................................................................... 57
Sorin C. Popescu
4. Impulse Synthetic Aperture Radar.................................................................................... 85
Giorgio Franceschetti and James Z. Tatoian
5. Hyperspectral Remote Sensing of Vegetation Bioparameters................................... 101
Ruiliang Pu and Peng Gong
6. Thermal Remote Sensing of Urban Areas: Theoretical Backgrounds
and Case Studies.................................................................................................................. 143
Qihao Weng

Section II Algorithms and Techniques
7. Atmospheric Correction Methods for Optical Remote Sensing
Imagery of Land................................................................................................................... 161
Rudolf Richter



8. Three-Dimensional Geometric Correction of Earth Observation
Satellite Data........................................................................................................................ 173
Thierry Toutin
9. Remote Sensing Image Classification............................................................................. 219
Dengsheng Lu, Qihao Weng, Emilio Moran, Guiying Li, and Scott Hetrick
10. Object-Based Image Analysis for Vegetation Mapping and Monitoring................ 241
Thomas Blaschke, Kasper Johansen, and Dirk Tiede
11. Land-Use and Land-Cover Change Detection............................................................... 273
Dengsheng Lu, Emilio Moran, Scott Hetrick, and Guiying Li


Section III Environmental Applications -Vegetation
12. Remote Sensing of Ecosystem Structure and Function............................................... 291
Alfredo R. Huete and Edward P. Glenn
13. Remote Sensing of Live Fuel Moisture........................................................................... 321
Stephen R. Yool
14. Forest Change Analysis Using Time-Series Landsat Observations......................... 339
Chengquan Huang
15. Satellite-Based Modeling of Gross Primary Production of
Terrestrial Ecosystems........................................................................................................ 367
Xiangming Xiao, Huimin Yan, Joshua Kalfas, and Qingyuan Zhang



16. Global Croplands and Their Water Use from Remote Sensing and
Nonremote Sensing Perspectives..................................................................................... 383
Prasad S. Thenkabail, Munir A. Hanjra, Venkateswarlu Dheeravath,
and Muralikrishna Gumma

Section IV Environmental Applications: Air, Water, and Land
17. Remote Sensing of Aerosols from Space: A Review of Aerosol Retrieval
Using the Moderate-Resolution Imaging Spectroradiometer.................................... 423
Man Sing Wong and Janet Nichol
18. Remote Estimation of Chlorophyll-a Concentration in Inland,
Estuarine, and Coastal Waters.......................................................................................... 439
Anatoly A. Gitelson, Daniela Gurlin, Wesley J. Moses, and Yosef Z. Yacobi
19. Retrievals of Turbulent Heat Fluxes and Surface Soil Water Content by
Remote Sensing................................................................................................................... 469
George P. Petropoulos and Toby N. Carlson



20. Remote Sensing of Urban Biophysical Environments................................................. 503
Qihao Weng
21. Development of the USGS National Land-Cover Database over Two Decades..... 525
George Xian, Collin Homer, and Limin Yang
Index............................................................................................................................................. 545

Download Here

Advances in Environmental Remote Sensing