Remote Sensing

7599 KiB

Open AccessArticle

Mapping Vegetation Density in a Heterogeneous River Floodplain Ecosystem Using Pointable CHRIS/PROBA Data

by Jochem Verrelst, Erika Romijn and Lammert Kooistra

Remote Sens. 2012, 4(9), 2866-2889; https://doi.org/10.3390/rs4092866 - 24 Sep 2012

Cited by 113 | Viewed by 11861

River floodplains in the Netherlands serve as water storage areas, while they also have the function of nature rehabilitation areas. Floodplain vegetation is therefore subject to natural processes of vegetation succession. At the same time, vegetation encroachment obstructs the water flow into the [...] Read more.

River floodplains in the Netherlands serve as water storage areas, while they also have the function of nature rehabilitation areas. Floodplain vegetation is therefore subject to natural processes of vegetation succession. At the same time, vegetation encroachment obstructs the water flow into the floodplains and increases the flood risk for the hinterland. Spaceborne pointable imaging spectroscopy has the potential to quantify vegetation density on the basis of leaf area index (LAI) from a desired view zenith angle. In this respect, hyperspectral pointable CHRIS data were linked to the ray tracing canopy reflectance model FLIGHT to retrieve vegetation density estimates over a heterogeneous river floodplain. FLIGHT enables simulating top-of-canopy reflectance of vegetated surfaces either in turbid (e.g., grasslands) or in 3D (e.g., forests) mode. By inverting FLIGHT against CHRIS data, LAI was computed for three main classified vegetation types, ‘herbaceous’, ‘shrubs’ and ‘forest’, and for the CHRIS view zenith angles in nadir, backward (−36°) and forward (+36°) scatter direction. The −36° direction showed most LAI variability within the vegetation types and was best validated, closely followed by the nadir direction. The +36° direction led to poorest LAI retrievals. The class-based inversion process has been implemented into a GUI toolbox which would enable the river manager to generate LAI maps in a semiautomatic way. Full article

► Show Figures

The study area which is located in the east of the Netherlands, indicated on the CHRIS nadir image in true colour band composition (R: 675.2 nm, G: 551.7 nm, B: 490.5 nm). The red circle represents the river floodplains of Millingerwaard. The black outlined river area overlain on the CHRIS nadir image represents the nature reserve the Gelderse Poort. Full article ">
Polar plot showing the actual positions of the five angular CHRIS images during acquisition on 6 September 2005. The solar zenith angle was 46°, the solar azimuth angle 170°. Full article ">
Maximum likelihood classification result of the CHRIS nadir image of the (a) Gelderse Poort and (b) Millingerwaard (indicated with the black square) into major land cover types. Full article ">
LAI maps (left) and derived histograms for LAI<8.5 (right) of Millingerwaard for the backward scattering direction (−36° VZA) (top), the nadir direction (middle) and the forward scattering direction (+36° VZA) (down), derived with FLIGHT model inversion. Full article ">
Mean validation results and standard deviation of the estimated LAI obtained with FLIGHT model inversion, plotted against the measured LAI values obtained with the hemispherical camera for the backward scattering direction (−36° VZA), the nadir direction and the forward scattering direction (+36° VZA). Full article ">
Maps of minimum RMSEs for LAI retrievals (left) and derived histograms for <8.5 (right) of Millingerwaard for the backward scattering direction (−36° VZA) (top), the nadir direction (middle) and the forward scattering direction (+36° VZA) (down), derived with FLIGHT model inversion. Full article ">
LAI map and histogram for the backward scattering direction (−36° VZA), derived with FLIGHT model inversion after applying to the Gelderse Poort area. Full article ">

734 KiB

Open AccessArticle

Multivariate Analysis of MODerate Resolution Imaging Spectroradiometer (MODIS) Aerosol Retrievals and the Statistical Hurricane Intensity Prediction Scheme (SHIPS) Parameters for Atlantic Hurricanes

by Mohammed M. Kamal, Ruixin Yang and John J. Qu

Remote Sens. 2012, 4(9), 2846-2865; https://doi.org/10.3390/rs4092846 - 24 Sep 2012

Cited by 1 | Viewed by 7228

Abstract

MODerate Resolution Imaging Spectroradiometer (MODIS) aerosol retrievals over the North Atlantic spanning seven hurricane seasons are combined with the Statistical Hurricane Intensity Prediction Scheme (SHIPS) parameters. The difference between the current and future intensity changes were selected as response variables. For 24 major [...] Read more.

MODerate Resolution Imaging Spectroradiometer (MODIS) aerosol retrievals over the North Atlantic spanning seven hurricane seasons are combined with the Statistical Hurricane Intensity Prediction Scheme (SHIPS) parameters. The difference between the current and future intensity changes were selected as response variables. For 24 major hurricanes (category 3, 4 and 5) between 2003 and 2009, eight lead time response variables were determined to be between 6 and 48 h. By combining MODIS and SHIPS data, 56 variables were compiled and selected as predictors for this study. Variable reduction from 56 to 31 was performed in two steps; the first step was via correlation coefficients (cc) followed by Principal Component Analysis (PCA) extraction techniques. The PCA reduced 31 variables to 20. Five categories were established based on the PCA group variables exhibiting similar physical phenomena. Average aerosol retrievals from MODIS Level 2 data in the vicinity of UTC 1,200 and 1,800 h were mapped to the SHIPS parameters to perform Multiple Linear Regression (MLR) between each response variable against six sets of predictors of 31, 30, 28, 27, 23 and 20 variables. The deviation among the predictors Root Mean Square Error (RMSE) varied between 0.01 through 0.05 and, therefore, implied that reducing the number of variables did not change the core physical information. Even when the parameters are reduced from 56 to 20, the correlation values exhibit a stronger relationship between the response and predictors. Therefore, the same phenomena can be explained by the reduction of variables. Full article

► Show Figures

9284 KiB

Open AccessArticle

Modelling Forest α-Diversity and Floristic Composition — On the Added Value of LiDAR plus Hyperspectral Remote Sensing

by Benjamin F. Leutner, Björn Reineking, Jörg Müller, Martin Bachmann, Carl Beierkuhnlein, Stefan Dech and Martin Wegmann

Remote Sens. 2012, 4(9), 2818-2845; https://doi.org/10.3390/rs4092818 - 21 Sep 2012

Cited by 74 | Viewed by 12616

Abstract

The decline of biodiversity is one of the major current global issues. Still, there is a widespread lack of information about the spatial distribution of individual species and biodiversity as a whole. Remote sensing techniques are increasingly used for biodiversity monitoring and especially [...] Read more.

The decline of biodiversity is one of the major current global issues. Still, there is a widespread lack of information about the spatial distribution of individual species and biodiversity as a whole. Remote sensing techniques are increasingly used for biodiversity monitoring and especially the combination of LiDAR and hyperspectral data is expected to deliver valuable information. In this study spatial patterns of vascular plant community composition and α-diversity of a temperate montane forest in Germany were analysed for different forest strata. The predictive power of LiDAR (LiD) and hyperspectral (MNF) datasets alone and combined (MNF+LiD) was compared using random forest regression in a ten-fold cross-validation scheme that included feature selection and model tuning. The final models were used for spatial predictions. Species richness could be predicted with varying accuracy (R² = 0.26 to 0.55) depending on the forest layer. In contrast, community composition of the different layers, obtained by multivariate ordination, could in part be modelled with high accuracies for the first ordination axis (R² = 0.39 to 0.78), but poor accuracies for the second axis (R² ≤ 0.3). LiDAR variables were the best predictors for total species richness across all forest layers (R²_LiD = 0.3, R² _MNF = 0.08, R² _MNF+LiD = 0.2), while for community composition across all forest layers both hyperspectral and LiDAR predictors achieved similar performances (R²_LiD = 0.75, R²_MNF = 0.76, R² _MNF+LiD = 0.78). The improvement in R² was small (≤0.07)—if any—when using both LiDAR and hyperspectral data as compared to using only the best single predictor set. This study shows the high potential of LiDAR and hyperspectral data for plant biodiversity modelling, but also calls for a critical evaluation of the added value of combining both with respect to acquisition costs. Full article

(This article belongs to the Special Issue Remote Sensing of Biological Diversity)

► Show Figures

Graphical abstract

Graphical abstract
Full article ">
Bavarian Forest national park with bounding boxes of the two hyperspectral scenes. N: northern scene, S: southern scene. Underlying digital elevation model from SRTM3 v2.1 [<a href="#b50-remotesensing-04-02818" class="html-bibr">50</a>]. Projection: left: WGS84 UTM 33N; right: WGS84. Full article ">
Processing framework. Predictor variables were derived from LiDAR (LiD) and hyperspectral (MNF) data. Biodiversity indicators (SR,H′) and non-metric multidimensional scaling (NMDS) ordination scores of two NMDS axes were modelled as response variables (NMDS1,NMDS2). The procedure was repeated for herb, shrub and tree layer and their total, using LiD, MNF and MNF+LiD predictors both with and without feature selection. Full article ">
Bi-plot of site and species scores in NMDS space for herb (HL), shrub (SL), tree layer (TL) and their total (ALL). Species acronyms consist of the first three letters of genus and species respectively (see supplementary materials, Table S1 for decoding). Only species occurring in more than 20% of the plots are shown. Colour mapping depicts species richness. Full article ">
Kendall’s τ rank correlation for inter-predictor correlation. The frequency histograms are based on the upper triangular correlation matrix (diagonal removed). The LiDAR and MNF panels are based on correlation between all predictors within a predictor set, whereas the MNF+LiD panel shows only the correlation coefficients between the ten LiDAR and the MNF variables. Full article ">
Model assessment of ten-fold cross-validation of random forests with and without feature selection and with mtry optimization. Response variables: Species richness (SR), Shannon index (H′) and ordination scores (NMDS1, NMDS2). Predictor variables: LiDAR (LiD) and hyperspectral (MNF) predictor sets and their combination (MNF+LiD). Bars: cross-validated <math display="inline"> <mrow> <msubsup> <mrow> <mi>R</mi></mrow> <mrow> <mtext>CV</mtext></mrow> <mn>2</mn></msubsup></mrow></math> with error bars depicting the bootstrapped standard deviation of <math display="inline"> <mrow> <msubsup> <mrow> <mi>R</mi></mrow> <mrow> <mtext>CV</mtext></mrow> <mn>2</mn></msubsup></mrow></math>. Points show the corresponding mean OOB estimates <math display="inline"> <mrow> <msubsup> <mrow> <mi>R</mi></mrow> <mrow> <mtext>OOB</mtext></mrow> <mn>2</mn></msubsup></mrow></math> averaged over the ten cross-validation groups together with their standard deviation in error bars. Values of R2 below zero are artifacts of the calculation, should be interpreted as zero and are hence displayed as such. Full article ">
Predicted species richness of the shrub layer (SL) for the LiDAR (LiD) and hyperspectral (MNF) predictor sets and their combination (MNF+LiD). Upper panel: averaged predictions of ten cross-validation models and plot information used to train and evaluate the random forests. Background: color-infrared composite of the HyMap data. Lower panel: standard deviation of the predictions of the ten cross-validation models. Bar charts show the residuals between predicted and observed estimates in cross-validation. Positive residuals are due to an overestimation by the model and vice versa. Full article ">
Spatial prediction of position on the first NMDS axis of all vegetation layers combined (ALL) for the LiDAR (LiD) and hyperspectral (MNF) predictor sets and their combination (MNF+LiD). Upper panel: averaged predictions of ten cross-validation models and plot information used to train and evaluate the random forests. Background: color-infrared composite of the HyMap data. Lower panel: standard deviation of the predictions of the ten cross-validation models. Bar charts show the residuals between predicted and observed estimates in cross-validation. Positive residuals are due to an overestimation by the model and vice versa. Full article ">
Spatial prediction of position on the first NMDS axis of the herb layer (HL) for the LiDAR (LiD) and hyperspectral (MNF) predictor sets and their combination (MNF+LiD). Upper panel: averaged predictions of ten cross-validation models and plot information used to train and evaluate the random forests. Background: color-infrared composite of the HyMap data. Lower panel: standard deviation of the predictions of the ten cross-validation models. Bar charts show the residuals between predicted and observed estimates in cross-validation. Positive residuals are due to an overestimation by the model and vice versa. Full article ">

5047 KiB

Open AccessArticle

Operational Automatic Remote Sensing Image Understanding Systems: Beyond Geographic Object-Based and Object-Oriented Image Analysis (GEOBIA/GEOOIA). Part 2: Novel system Architecture, Information/Knowledge Representation, Algorithm Design and Implementation

by Andrea Baraldi and Luigi Boschetti

Remote Sens. 2012, 4(9), 2768-2817; https://doi.org/10.3390/rs4092768 - 20 Sep 2012

Cited by 17 | Viewed by 10043

Abstract

According to literature and despite their commercial success, state-of-the-art two-stage non-iterative geographic object-based image analysis (GEOBIA) systems and three-stage iterative geographic object-oriented image analysis (GEOOIA) systems, where GEOOIA/GEOBIA, remain affected by a lack of productivity, general consensus and research. To outperform the Quality [...] Read more.

According to literature and despite their commercial success, state-of-the-art two-stage non-iterative geographic object-based image analysis (GEOBIA) systems and three-stage iterative geographic object-oriented image analysis (GEOOIA) systems, where GEOOIA/GEOBIA, remain affected by a lack of productivity, general consensus and research. To outperform the Quality Indexes of Operativeness (OQIs) of existing GEOBIA/GEOOIA systems in compliance with the Quality Assurance Framework for Earth Observation (QA4EO) guidelines, this methodological work is split into two parts. Based on an original multi-disciplinary Strengths, Weaknesses, Opportunities and Threats (SWOT) analysis of the GEOBIA/GEOOIA approaches, the first part of this work promotes a shift of learning paradigm in the pre-attentive vision first stage of a remote sensing (RS) image understanding system (RS-IUS), from sub-symbolic statistical model-based (inductive) image segmentation to symbolic physical model-based (deductive) image preliminary classification capable of accomplishing image sub-symbolic segmentation and image symbolic pre-classification simultaneously. In the present second part of this work, a novel hybrid (combined deductive and inductive) RS-IUS architecture featuring a symbolic deductive pre-attentive vision first stage is proposed and discussed in terms of: (a) computational theory (system design), (b) information/knowledge representation, (c) algorithm design and (d) implementation. As proof-of-concept of symbolic physical model-based pre-attentive vision first stage, the spectral knowledge-based, operational, near real-time, multi-sensor, multi-resolution, application-independent Satellite Image Automatic Mapper™ (SIAM™) is selected from existing literature. To the best of these authors’ knowledge, this is the first time a symbolic syntactic inference system, like SIAM™, is made available to the RS community for operational use in a RS-IUS pre-attentive vision first stage, to accomplish multi-scale image segmentation and multi-granularity image pre-classification simultaneously, automatically and in near real-time. Full article

► Show Figures

Graphical abstract

3706 KiB

Open AccessArticle

Mapping of Ice Motion in Antarctica Using Synthetic-Aperture Radar Data

by Jeremie Mouginot, Bernd Scheuchl and Eric Rignot

Remote Sens. 2012, 4(9), 2753-2767; https://doi.org/10.3390/rs4092753 - 18 Sep 2012

Cited by 184 | Viewed by 13168

Abstract

Ice velocity is a fundamental parameter in studying the dynamics of ice sheets. Until recently, no complete mapping of Antarctic ice motion had been available due to calibration uncertainties and lack of basic data. Here, we present a method for calibrating and mosaicking [...] Read more.

Ice velocity is a fundamental parameter in studying the dynamics of ice sheets. Until recently, no complete mapping of Antarctic ice motion had been available due to calibration uncertainties and lack of basic data. Here, we present a method for calibrating and mosaicking an ensemble of InSAR satellite measurements of ice motion from six sensors: the Japanese ALOS PALSAR, the European Envisat ASAR, ERS-1 and ERS-2, and the Canadian RADARSAT-1 and RADARSAT-2. Ice motion calibration is made difficult by the sparsity of in-situ reference points and the shear size of the study area. A sensor-dependent data stacking scheme is applied to reduce measurement uncertainties. The resulting ice velocity mosaic has errors in magnitude ranging from 1 m/yr in the interior regions to 17 m/yr in coastal sectors and errors in flow direction ranging from less than 0.5° in areas of fast flow to unconstrained direction in sectors of slow motion. It is important to understand how these mosaics are calibrated to understand the inner characteristics of the velocity products as well as to plan future InSAR acquisitions in the Antarctic. As an example, we show that in broad sectors devoid of ice-motion control, it is critical to operate ice motion mapping on a large scale to avoid pitfalls of calibration uncertainties that would make it difficult to obtain quality products and especially construct reliable time series of ice motion needed to detect temporal changes. Full article

(This article belongs to the Special Issue Remote Sensing by Synthetic Aperture Radar Technology)

► Show Figures

2984 KiB

Open AccessArticle

Radiometric and Geometric Analysis of Hyperspectral Imagery Acquired from an Unmanned Aerial Vehicle

by Ryan Hruska, Jessica Mitchell, Matthew Anderson and Nancy F. Glenn

Remote Sens. 2012, 4(9), 2736-2752; https://doi.org/10.3390/rs4092736 - 17 Sep 2012

Cited by 163 | Viewed by 17129

Abstract

In the summer of 2010, an Unmanned Aerial Vehicle (UAV) hyperspectral calibration and characterization experiment of the Resonon PIKA II imaging spectrometer was conducted at the US Department of Energy’s Idaho National Laboratory (INL) UAV Research Park. The purpose of the experiment was [...] Read more.

In the summer of 2010, an Unmanned Aerial Vehicle (UAV) hyperspectral calibration and characterization experiment of the Resonon PIKA II imaging spectrometer was conducted at the US Department of Energy’s Idaho National Laboratory (INL) UAV Research Park. The purpose of the experiment was to validate the radiometric calibration of the spectrometer and determine the georegistration accuracy achievable from the on-board global positioning system (GPS) and inertial navigation sensors (INS) under operational conditions. In order for low-cost hyperspectral systems to compete with larger systems flown on manned aircraft, they must be able to collect data suitable for quantitative scientific analysis. The results of the in-flight calibration experiment indicate an absolute average agreement of 96.3%, 93.7% and 85.7% for calibration tarps of 56%, 24%, and 2.5% reflectivity, respectively. The achieved planimetric accuracy was 4.6 m (based on RMSE) with a flying height of 344 m above ground level (AGL). Full article

(This article belongs to the Special Issue Unmanned Aerial Vehicles (UAVs) based Remote Sensing)

► Show Figures

659 KiB

Open AccessArticle

Operational Automatic Remote Sensing Image Understanding Systems: Beyond Geographic Object-Based and Object-Oriented Image Analysis (GEOBIA/GEOOIA). Part 1: Introduction

by Andrea Baraldi and Luigi Boschetti

Remote Sens. 2012, 4(9), 2694-2735; https://doi.org/10.3390/rs4092694 - 14 Sep 2012

Cited by 39 | Viewed by 11112

Abstract

According to existing literature and despite their commercial success, state-of-the-art two-stage non-iterative geographic object-based image analysis (GEOBIA) systems and three-stage iterative geographic object-oriented image analysis (GEOOIA) systems, where GEOOIA/GEOBIA, remain affected by a lack of productivity, general consensus and research. To outperform the [...] Read more.

According to existing literature and despite their commercial success, state-of-the-art two-stage non-iterative geographic object-based image analysis (GEOBIA) systems and three-stage iterative geographic object-oriented image analysis (GEOOIA) systems, where GEOOIA/GEOBIA, remain affected by a lack of productivity, general consensus and research. To outperform the degree of automation, accuracy, efficiency, robustness, scalability and timeliness of existing GEOBIA/GEOOIA systems in compliance with the Quality Assurance Framework for Earth Observation (QA4EO) guidelines, this methodological work is split into two parts. The present first paper provides a multi-disciplinary Strengths, Weaknesses, Opportunities and Threats (SWOT) analysis of the GEOBIA/GEOOIA approaches that augments similar analyses proposed in recent years. In line with constraints stemming from human vision, this SWOT analysis promotes a shift of learning paradigm in the pre-attentive vision first stage of a remote sensing (RS) image understanding system (RS-IUS), from sub-symbolic statistical model-based (inductive) image segmentation to symbolic physical model-based (deductive) image preliminary classification. Hence, a symbolic deductive pre-attentive vision first stage accomplishes image sub-symbolic segmentation and image symbolic pre-classification simultaneously. In the second part of this work a novel hybrid (combined deductive and inductive) RS-IUS architecture featuring a symbolic deductive pre-attentive vision first stage is proposed and discussed in terms of: (a) computational theory (system design); (b) information/knowledge representation; (c) algorithm design; and (d) implementation. As proof-of-concept of symbolic physical model-based pre-attentive vision first stage, the spectral knowledge-based, operational, near real-time Satellite Image Automatic Mapper™ (SIAM™) is selected from existing literature. To the best of these authors’ knowledge, this is the first time a symbolic syntactic inference system, like SIAM™, is made available to the RS community for operational use in a RS-IUS pre-attentive vision first stage, to accomplish multi-scale image segmentation and multi-granularity image pre-classification simultaneously, automatically and in near real-time. Full article

► Show Figures

Graphical abstract

1286 KiB

Open AccessArticle

Tree Species Classification with Random Forest Using Very High Spatial Resolution 8-Band WorldView-2 Satellite Data

by Markus Immitzer, Clement Atzberger and Tatjana Koukal

Remote Sens. 2012, 4(9), 2661-2693; https://doi.org/10.3390/rs4092661 - 14 Sep 2012

Cited by 666 | Viewed by 37047

Abstract

Tree species diversity is a key parameter to describe forest ecosystems. It is, for example, important for issues such as wildlife habitat modeling and close-to-nature forest management. We examined the suitability of 8-band WorldView-2 satellite data for the identification of 10 tree species [...] Read more.

Tree species diversity is a key parameter to describe forest ecosystems. It is, for example, important for issues such as wildlife habitat modeling and close-to-nature forest management. We examined the suitability of 8-band WorldView-2 satellite data for the identification of 10 tree species in a temperate forest in Austria. We performed a Random Forest (RF) classification (object-based and pixel-based) using spectra of manually delineated sunlit regions of tree crowns. The overall accuracy for classifying 10 tree species was around 82% (8 bands, object-based). The class-specific producer’s accuracies ranged between 33% (European hornbeam) and 94% (European beech) and the user’s accuracies between 57% (European hornbeam) and 92% (Lawson’s cypress). The object-based approach outperformed the pixel-based approach. We could show that the 4 new WorldView-2 bands (Coastal, Yellow, Red Edge, and Near Infrared 2) have only limited impact on classification accuracy if only the 4 main tree species (Norway spruce, Scots pine, European beech, and English oak) are to be separated. However, classification accuracy increased significantly using the full spectral resolution if further tree species were included. Beside the impact on overall classification accuracy, the importance of the spectral bands was evaluated with two measures provided by RF. An in-depth analysis of the RF output was carried out to evaluate the impact of reference data quality and the resulting reliability of final class assignments. Finally, an extensive literature review on tree species classification comprising about 20 studies is presented. Full article

(This article belongs to the Special Issue Remote Sensing of Biological Diversity)

► Show Figures

Graphical abstract

1774 KiB

Open AccessArticle

Flux Measurements in Cairo. Part 2: On the Determination of the Spatial Radiation and Energy Balance Using ASTER Satellite Data

by Corinne Myrtha Frey and Eberhard Parlow

Remote Sens. 2012, 4(9), 2635-2660; https://doi.org/10.3390/rs4092635 - 13 Sep 2012

Cited by 21 | Viewed by 7290

Abstract

This study highlights the possibilities and constraints of determining instantaneous spatial surface radiation and land heat fluxes from satellite images in a heterogeneous urban area and its agricultural and natural surroundings. Net radiation was determined using ASTER satellite data and MODTRAN radiative transfer [...] Read more.

This study highlights the possibilities and constraints of determining instantaneous spatial surface radiation and land heat fluxes from satellite images in a heterogeneous urban area and its agricultural and natural surroundings. Net radiation was determined using ASTER satellite data and MODTRAN radiative transfer calculations. The soil heat flux was estimated with two empirical methods using radiative terms and vegetation indices. The turbulent heat fluxes finally were determined with the LUMPS (Local-Scale Urban Meteorological Parameterization Scheme) and the ARM (Aerodynamic Resistance Method) method. Results were compared to in situ measured ground data. The performance of the atmospheric correction was found to be crucial for the estimation of the radiation balance and thereafter the heat fluxes. The soil heat flux could be modeled satisfactorily by both of the applied approaches. The LUMPS method, for the turbulent fluxes, appeals by its simplicity. However, a correct spatial estimation of associated parameters could not always be achieved. The ARM method showed the better spatial results for the turbulent heat fluxes. In comparison with the in situ measurements however, the LUMPS approach rendered the better results than the ARM method. Full article

(This article belongs to the Special Issue Thermal Remote Sensing Applications: Present Status and Future Possibilities)

► Show Figures

1418 KiB

Open AccessArticle

Remote Sensing of Fractional Green Vegetation Cover Using Spatially-Interpolated Endmembers

by Brian Johnson, Ryutaro Tateishi and Toshiyuki Kobayashi

Remote Sens. 2012, 4(9), 2619-2634; https://doi.org/10.3390/rs4092619 - 12 Sep 2012

Cited by 60 | Viewed by 9206

Abstract

Fractional green vegetation cover (FVC) is a useful parameter for many environmental and climate-related applications. A common approach for estimating FVC involves the linear unmixing of two spectral endmembers in a remote sensing image; bare soil and green vegetation. The spectral properties of [...] Read more.

Fractional green vegetation cover (FVC) is a useful parameter for many environmental and climate-related applications. A common approach for estimating FVC involves the linear unmixing of two spectral endmembers in a remote sensing image; bare soil and green vegetation. The spectral properties of these two endmembers are typically determined based on field measurements, estimated using additional data sources (e.g., soil databases or land cover maps), or extracted directly from the imagery. Most FVC estimation approaches do not consider that the spectral properties of endmembers may vary across space. However, due to local differences in climate, soil type, vegetation species, etc., the spectral characteristics of soil and green vegetation may exhibit positive spatial autocorrelation. When this is the case, it may be useful to take these local variations into account for estimating FVC. In this study, spatial interpolation (Inverse Distance Weighting and Ordinary Kriging) was used to predict variations in the spectral characteristics of bare soil and green vegetation across space. When the spatially-interpolated values were used in place of scene-invariant endmember values to estimate FVC in an Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) image, the accuracy of FVC estimates increased, providing evidence that it may be useful to consider the effects of spatial autocorrelation for spectral mixture analysis. Full article

► Show Figures

Graphical abstract

Graphical abstract
Full article ">
False color Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) image, with the study area outlined in black (centered at 36°57′N, 140°38′E). NIR, R, and Green ASTER bands are shown in red, green, and blue color, respectively. Full article ">
Normalized Differential Vegetation Index (NDVI) (a) and Modified Soil Adjusted Vegetation Index (MSAVI) (b) images of the study area. The topographic effects evident in <a href="#f1-remotesensing-04-02619" class="html-fig">Figure 1</a> have been largely removed using the vegetation indices. Full article ">
Predicted NDVIs (a) and NDVIv (b) values using Inverse Distance Weighting (IDW) interpolation. Yellow points show the locations of gv_0 (a) and gv_1 (b) samples. Optimal IDW exponent value was 1.12 for (a) and 2.45 for (b). Full article ">
Predicted NDVIs (a) and NDVIv (b) values using Ordinary Kriging (OK) interpolation. Yellow points show the locations of gv_0 (a) and gv_1 (b) samples. Full article ">
Scatterplots of Reference and Estimated FVC using the NDVI (a), OK-NDVI (b), MSAVI (c), and OK-MSAVI (d) approach. OK-NDVI and OK-MSAVI show a modestly better linear fit than NDVI and MSAVI (based on R2 values). Full article ">
Scatterplots of Reference and Estimated FVC using the NDVI (a), OK-NDVI (b), MSAVI (c), and OK-MSAVI (d) approach. OK-NDVI and OK-MSAVI show a modestly better linear fit than NDVI and MSAVI (based on R2 values). Full article ">
Estimated FVC (%) using the most accurate estimation method, OK-NDVI. Full article ">

4139 KiB

Open AccessArticle

Overcoming Limitations with Landsat Imagery for Mapping of Peat Swamp Forests in Sundaland

by Lahiru S. Wijedasa, Sean Sloan, Dimitrios G. Michelakis and Gopalasamy R. Clements

Remote Sens. 2012, 4(9), 2595-2618; https://doi.org/10.3390/rs4092595 - 10 Sep 2012

Cited by 57 | Viewed by 15539

Abstract

Landsat can be used to map tropical forest cover at 15–60 m resolution, which is helpful for detecting small but important perturbations in increasingly fragmented forests. However, among the remaining Landsat satellites, Landsat-5 no longer has global coverage and, since 2003, a mechanical [...] Read more.

Landsat can be used to map tropical forest cover at 15–60 m resolution, which is helpful for detecting small but important perturbations in increasingly fragmented forests. However, among the remaining Landsat satellites, Landsat-5 no longer has global coverage and, since 2003, a mechanical fault in the Scan-Line Corrector (SLC-Off) of the Landsat-7 satellite resulted in a 22–25% data loss in each image. Such issues challenge the use of Landsat for wall-to-wall mapping of tropical forests, and encourage the use of alternative, spatially coarser imagery such as MODIS. Here, we describe and test an alternative method of post-classification compositing of Landsat images for mapping over 20.5 million hectares of peat swamp forest in the biodiversity hotspot of Sundaland. In order to reduce missing data to levels comparable to those prior to the SLC-Off error, we found that, for a combination of Landsat-5 images and SLC-off Landsat-7 images used to create a 2005 composite, 86% of the 58 scenes required one or two images, while 14% required three or more images. For a 2010 composite made using only SLC-Off Landsat-7 images, 64% of the scenes required one or two images and 36% required four or more images. Missing-data levels due to cloud cover and shadows in the pre SLC-Off composites (7.8% and 10.3% for 1990 and 2000 enhanced GeoCover mosaics) are comparable to the post SLC-Off composites (8.2% and 8.3% in the 2005 and 2010 composites). The area-weighted producer’s accuracy for our 2000, 2005 and 2010 composites were 77%, 85% and 86% respectively. Overall, these results show that missing-data levels, classification accuracy, and geographic coverage of Landsat composites are comparable across a 20-year period despite the SLC-Off error since 2003. Correspondingly, Landsat still provides an appreciable utility for monitoring tropical forests, particularly in Sundaland’s rapidly disappearing peat swamp forests. Full article

(This article belongs to the Special Issue Remote Sensing of Biological Diversity)

► Show Figures

Graphical abstract

701 KiB

Open AccessArticle

Discrimination of Switchgrass Cultivars and Nitrogen Treatments Using Pigment Profiles and Hyperspectral Leaf Reflectance Data

by Anserd J. Foster, Vijaya Gopal Kakani, Jianjun Ge and Jagadeesh Mosali

Remote Sens. 2012, 4(9), 2576-2594; https://doi.org/10.3390/rs4092576 - 10 Sep 2012

Cited by 15 | Viewed by 7819

Abstract

The objective of this study was to compare the use of hyperspectral narrowbands, hyperspectral narrowband indices and pigment measurements collected from switchgrass leaf as potential tools for discriminating among twelve switchgrass cultivars and five N treatments in one cultivar (Alamo). Hyperspectral reflectance, UV-B [...] Read more.

The objective of this study was to compare the use of hyperspectral narrowbands, hyperspectral narrowband indices and pigment measurements collected from switchgrass leaf as potential tools for discriminating among twelve switchgrass cultivars and five N treatments in one cultivar (Alamo). Hyperspectral reflectance, UV-B absorbing compounds, photosynthetic pigments (chlorophyll a, chlorophyll b and carotenoids) of the uppermost fully expanded leaves were determined at monthly intervals from May to September. Leaf hyperspectral data was collected using ASD FieldSpec FR spectroradiometer (350–2,500 nm). Discrimination of the cultivars and N treatments were determined based on Principal Component Analysis (PCA) and linear discriminant analysis (DA). The stepwise discriminant analysis was used to determine the best indices that differentiate switchgrass cultivars and nitrogen treatments. Results of PCA showed 62% of the variability could be explained in PC1 dominated by middle infrared wavebands, over 20% in PC2 dominated by near infrared wavebands and just over 10% in PC3 dominated by green wavebands for separating both cultivars and N treatments. Discriminating among the cultivars resulted in an overall accuracy of 81% with the first five PCs in the month of September, but was less accurate (27%) in classifying N treatments using the spectral data. Discrimination based on pigment data using the first two PCs resulted in an overall accuracy of less than 10% for separating switchgrass cultivars , but was more accurate (47%) in grouping N treatments. The plant senescence ratio index (PSRI) was found to be the best index for separating the cultivars late in the season, while the transform chlorophyll absorption ration index (TCARI) was best for separating the N treatments. Leaf spectra data was found to be more useful than pigment data for the discrimination of switchgrass cultivars, particularly late in the growing season. Full article

► Show Figures

37123 KiB

Open AccessArticle

Estimation of Supraglacial Dust and Debris Geochemical Composition via Satellite Reflectance and Emissivity

by Kimberly Casey and Andreas Kääb

Remote Sens. 2012, 4(9), 2554-2575; https://doi.org/10.3390/rs4092554 - 7 Sep 2012

Cited by 10 | Viewed by 9264

Abstract

We demonstrate spectral estimation of supraglacial dust, debris, ash and tephra geochemical composition from glaciers and ice fields in Iceland, Nepal, New Zealand and Switzerland. Surface glacier material was collected and analyzed via X-ray fluorescence spectroscopy (XRF) and X-ray diffraction (XRD) for geochemical [...] Read more.

We demonstrate spectral estimation of supraglacial dust, debris, ash and tephra geochemical composition from glaciers and ice fields in Iceland, Nepal, New Zealand and Switzerland. Surface glacier material was collected and analyzed via X-ray fluorescence spectroscopy (XRF) and X-ray diffraction (XRD) for geochemical composition and mineralogy. In situ data was used as ground truth for comparison with satellite derived geochemical results. Supraglacial debris spectral response patterns and emissivity-derived silica weight percent are presented. Qualitative spectral response patterns agreed well with XRF elemental abundances. Quantitative emissivity estimates of supraglacial SiO2 in continental areas were 67% (Switzerland) and 68% (Nepal), while volcanic supraglacial SiO2 averages were 58% (Iceland) and 56% (New Zealand), yielding general agreement. Ablation season supraglacial temperature variation due to differing dust and debris type and coverage was also investigated, with surface debris temperatures ranging from 5.9 to 26.6 °C in the study regions. Applications of the supraglacial geochemical reflective and emissive characterization methods include glacier areal extent mapping, debris source identification, glacier kinematics and glacier energy balance considerations. Full article

► Show Figures

Graphical abstract

Graphical abstract
Full article ">
Map of study region locations, indicated by the cyan boxes. Full article ">
ASTER (band 3,2,1) images of Zmutt glacier (top left, acquired 29 July 2004) and Khumbu glacier (top right, acquired 29 November 2005). VNIR-SWIR multispectral AST_07XT reflectance is plotted beneath the images. Zmutt glacier serpentine Fe rich and Fe poor longitudinal debris bands can be differentiated from the Khumbu glacier leucogranite lighter granitic and darker schistic longitudinal debris bands. Full article ">
Images of the Iceland study region (top left, MODIS bands 1,4,3, acquired 28 August 2010) and Mt. Ruapehu, New Zealand study region (top right, ASTER bands 3,2,1, acquired 9 January 2008) are shown below. VNIR-SWIR multispectral MOD09GA (Iceland) and AST_07XT (New Zealand) surface reflectance means are plotted beneath the images. Note, the red box on Hofsjökull indicates the coverage of the Hyperion scene, discussed in Section 4.2, <a href="#f4-remotesensing-04-02554" class="html-fig">Figure 4</a>. Full article ">
Hofsjökull Hyperion (24 September 2001) at-sensor top-of-atmosphere supraglacial reflectance plotted at left, and the true color image is displayed on the right. Circles on the Hofsjökull image display the areas with varying tephra coverage in which reflectance data was selected to plot. The same colors are used for circles as the top-of-atmosphere reflectance response patterns (e.g., lightest orange indicating high firn, black indicating heaviest tephra covered ice). Atmospheric water vapor, oxygen and CO2 absorption features are removed. Full article ">
Shortwave and thermal infrared false color image composites of study regions differentiate silica rich (yellow coloring) vs. silica poor (blue coloring) supraglacial debris mineralogy. Full article ">
Lhotse Shar (reflectance data taken at red dot) and Imja glacier (reflectance data taken at blue dot) Hyperion at-sensor top-of-atmosphere reflectance of supraglacial ice and debris, taken from 13 May 2002 Hyperion acquisition. Atmospheric water vapor, oxygen and CO2 absorption features were removed. The true color image composite is displayed on the right, and was acquired by ALI on 4 October 2010 (10m pan enhanced image shown). The black “X” on upper Lhotse Shar glacier shows the area used for the spectral reflectance data collection. Full article ">

2653 KiB

Open AccessArticle

Land Cover and Land Use Classification with TWOPAC: towards Automated Processing for Pixel- and Object-Based Image Classification

by Juliane Huth, Claudia Kuenzer, Thilo Wehrmann, Steffen Gebhardt, Vo Quoc Tuan and Stefan Dech

Remote Sens. 2012, 4(9), 2530-2553; https://doi.org/10.3390/rs4092530 - 7 Sep 2012

Cited by 58 | Viewed by 15135

Abstract

We present a novel and innovative automated processing environment for the derivation of land cover (LC) and land use (LU) information. This processing framework named TWOPAC (TWinned Object and Pixel based Automated classification Chain) enables the standardized, independent, user-friendly, and comparable derivation of [...] Read more.

We present a novel and innovative automated processing environment for the derivation of land cover (LC) and land use (LU) information. This processing framework named TWOPAC (TWinned Object and Pixel based Automated classification Chain) enables the standardized, independent, user-friendly, and comparable derivation of LC and LU information, with minimized manual classification labor. TWOPAC allows classification of multi-spectral and multi-temporal remote sensing imagery from different sensor types. TWOPAC enables not only pixel-based classification, but also allows classification based on object-based characteristics. Classification is based on a Decision Tree approach (DT) for which the well-known C5.0 code has been implemented, which builds decision trees based on the concept of information entropy. TWOPAC enables automatic generation of the decision tree classifier based on a C5.0-retrieved ascii-file, as well as fully automatic validation of the classification output via sample based accuracy assessment.Envisaging the automated generation of standardized land cover products, as well as area-wide classification of large amounts of data in preferably a short processing time, standardized interfaces for process control, Web Processing Services (WPS), as introduced by the Open Geospatial Consortium (OGC), are utilized. TWOPAC’s functionality to process geospatial raster or vector data via web resources (server, network) enables TWOPAC’s usability independent of any commercial client or desktop software and allows for large scale data processing on servers. Furthermore, the components of TWOPAC were built-up using open source code components and are implemented as a plug-in for Quantum GIS software for easy handling of the classification process from the user’s perspective. Full article

► Show Figures

Flowchart of the Twinned Object- and Pixel based Automated classification Chain (TWOPAC). Full article ">
Sampling procedure—selection and extraction of segments with features. Full article ">
Classification result of TWOPAC run with Rapid Eye mosaic (2010-01-27) from the central part of the Mekong Delta—subset of object-based result (a); detail on rural area and natural tree area (b); detail on urban area (c). Full article ">
Classification result of TWOPAC run with SPOT5 dataset (2008-01-08) from the western coast of the Mekong Delta—subset of pixel-based result ((a) detail on rural area and natural tree area in (b); detail on urban area (c)); subset of object-based result ((d) details in (e,f)). Full article ">
Classification result of TWOPAC run with SPOT 4 dataset (2010-01-02) from the Dongting lake, Hunan province, China: pixel-based result generated with C5.0 ((a), details in (b,c)); for comparison maximum likelihood classification ((d), details in (e,f)). Full article ">
Classification result of TWOPAC run with multi-temporal MODIS 8 tiles mosaic (2009) for Central Asia: Overview of the classification result for Uzbekistan (a); details on specific land cover types—cultivated area and natural vegetation (b,c). Full article ">

1757 KiB

Open AccessArticle

Hyperspectral Time Series Analysis of Native and Invasive Species in Hawaiian Rainforests

by Ben Somers and Gregory P. Asner

Remote Sens. 2012, 4(9), 2510-2529; https://doi.org/10.3390/rs4092510 - 29 Aug 2012

Cited by 57 | Viewed by 10685

Abstract

The unique ecosystems of the Hawaiian Islands are progressively being threatened following the introduction of exotic species. Operational implementation of remote sensing for the detection, mapping and monitoring of these biological invasions is currently hampered by a lack of knowledge on the spectral [...] Read more.

The unique ecosystems of the Hawaiian Islands are progressively being threatened following the introduction of exotic species. Operational implementation of remote sensing for the detection, mapping and monitoring of these biological invasions is currently hampered by a lack of knowledge on the spectral separability between native and invasive species. We used spaceborne imaging spectroscopy to analyze the seasonal dynamics of the canopy hyperspectral reflectance properties of four tree species: (i) Metrosideros polymorpha, a keystone native Hawaiian species; (ii) Acacia koa, a native Hawaiian nitrogen fixer; (iii) the highly invasive Psidium cattleianum; and (iv) Morella faya, a highly invasive nitrogen fixer. The species specific separability of the reflectance and derivative-reflectance signatures extracted from an Earth Observing-1 Hyperion time series, composed of 22 cloud-free images spanning a period of four years and was quantitatively evaluated using the Separability Index (SI). The analysis revealed that the Hawaiian native trees were universally unique from the invasive trees in their near-infrared-1 (700–1,250 nm) reflectance (0.4 > SI > 1.4). Due to its higher leaf area index, invasive trees generally had a higher near-infrared reflectance. To a lesser extent, it could also be demonstrated that nitrogen-fixing trees were spectrally unique from non-fixing trees. The higher leaf nitrogen content of nitrogen-fixing trees was expressed through slightly increased separabilities in visible and shortwave-infrared reflectance wavebands (SI = 0.4). We also found phenology to be key to spectral separability analysis. As such, it was shown that the spectral separability in the near-infrared-1 reflectance between the native and invasive species groups was more expressed in summer (SI > 0.7) than in winter (SI < 0.7). The lowest separability was observed for March-July (SI < 0.3). This could be explained by the invasives taking advantage of the warmer summer period to expand their canopy. There was, however, no specific time window or a single spectral region that always defined the separability of all species groups, and thus intensive monitoring of plant phenology as well as the use of the full-range (400–2,500 nm) spectrum was highly advantageous in differentiating each species. These results set a basis for an operational invasive species monitoring program in Hawai’i using spaceborne imaging spectroscopy. Full article

(This article belongs to the Special Issue Remote Sensing of Biological Diversity)

► Show Figures

897 KiB

Open AccessArticle

Monitoring Biennial Bearing Effect on Coffee Yield Using MODIS Remote Sensing Imagery

by Tiago Bernardes, Maurício Alves Moreira, Marcos Adami, Angélica Giarolla and Bernardo Friedrich Theodor Rudorff

Remote Sens. 2012, 4(9), 2492-2509; https://doi.org/10.3390/rs4092492 - 27 Aug 2012

Cited by 72 | Viewed by 11918

Abstract

Coffee is the second most valuable traded commodity worldwide. Brazil is the world’s largest coffee producer, responsible for one third of the world production. A coffee plot exhibits high and low production in alternated years, a characteristic so called biennial yield. High yield [...] Read more.

Coffee is the second most valuable traded commodity worldwide. Brazil is the world’s largest coffee producer, responsible for one third of the world production. A coffee plot exhibits high and low production in alternated years, a characteristic so called biennial yield. High yield is generally a result of suitable conditions of foliar biomass. Moreover, in high production years one plot tends to lose more leaves than it does in low production years. In both cases some correlation between coffee yield and leaf biomass can be deduced which can be monitored through time series of vegetation indices derived from satellite imagery. In Brazil, a comprehensive, spatially distributed study assessing this relationship has not yet been done. The objective of this study was to assess possible correlations between coffee yield and MODIS derived vegetation indices in the Brazilian largest coffee-exporting province. We assessed EVI and NDVI MODIS products over the period between 2002 and 2009 in the south of Minas Gerais State whose production accounts for about one third of the Brazilian coffee production. Landsat images were used to obtain a reference map of coffee areas and to identify MODIS 250 m pure pixels overlapping homogeneous coffee crops. Only MODIS pixels with 100% coffee were included in the analysis. A wavelet-based filter was used to smooth EVI and NDVI time profiles. Correlations were observed between variations on yield of coffee plots and variations on vegetation indices for pixels overlapping the same coffee plots. The vegetation index metrics best correlated to yield were the amplitude and the minimum values over the growing season. The best correlations were obtained between variation on yield and variation on vegetation indices the previous year (R = 0.74 for minEVI metric and R = 0.68 for minNDVI metric). Although correlations were not enough to estimate coffee yield exclusively from vegetation indices, trends properly reflect the biennial bearing effect on coffee yield. Full article

(This article belongs to the Special Issue Advances in Remote Sensing of Agriculture)

► Show Figures

The study area in the south of Minas Gerais State. Full article ">
Overlap of the limits of MODIS pixels with coffee fields in TM/Landsat 5 image, false color composite color 3B4R5G, (A) pixels fully occupied by coffee crop and (B) crop variability within a MODIS pixel in Landsat images. Full article ">
Flowchart of data processing adopted in the study. Full article ">
Annual variation of vegetation indices for the selected pixels (A). Standard crop with the maximum vegetation index value for March (B) and minimum vegetation index for August (C) on Image TM/Landsat 3B4R5G. Full article ">
Filtered EVI and NDVI time series for coffee crop and the original data (without filtering process). Full article ">
Correlation between variation on coffee yield and variation on minimum values of vegetation indices (minEVI and minNDVI) for the same year. Full article ">
Correlation between variation on minimum values of vegetation indices (minEVI and minNDVI) and variation on coffee yield the following year. Full article ">
Water balance for Guaxupé location and Pearson correlation coefficients from 2002 to 2009. Full article ">

Journal Menu

Journal Browser

Remote Sens., Volume 4, Issue 9 (September 2012) – 16 articles , Pages 2492-2889

Further Information

Guidelines

MDPI Initiatives

Follow MDPI