Following a site-specific secondary succession in the Amazon using the Landsat CDR product and field inventory data

Secondary forests cover large areas and are strong carbon sinks in tropical regions. They are important for ecosystem functioning, biodiversity conservation, watershed protection, and recovery of soil fertility. In this study, we used the Surface Reflectance Climate Data Record (CDR) product from 16 Thematic Mapper (TM)/Landsat-5 images (1984–2010) to continuously track the secondary succession (SS) of a forest following land abandonment in 1980. Changes in canopy structure and floristic composition were analysed using data from four field inventories (1995, 2002, 2007, and 2012). To characterize variations in brightness, greenness, spectral reflectance, and shadows with the natural regeneration of vegetation, we applied tasselled cap transformations, principal component analysis (PCA), and linear spectral mixture models to the TM datasets. Shade fractions were plotted over time and correlated with the enhanced vegetation index (EVI) and the normalized difference vegetation index (NDVI). Because image texture may reflect the variability of the successional process, eight co-occurrence-based filter metrics were calculated for selected TM bands and plotted as a function of time since abandonment. The successional forest was compared to a nearby primary reference forest (PF) and had differences in the spectral and textural means evaluated using analysis of variance (ANOVA). The results showed increases of 35% and 10.4% over time in basal area and tree height, respectively. Species richness within the assemblage of sampling units increased from 14 to 71 between 1995 and 2012, and this trend was also confirmed using an individual-based rarefaction analysis. Species richness in 2012 was still lower than that observed in the PF site, which presented greater amounts of aboveground biomass (336.4 ± 17.0 ton ha^?1 for PF versus 98.5 ± 21.4 ton ha^?1 for SS in 2012). Brightness and greenness tasselled cap differences between the SS and PF rapidly decreased from 1984 (SS at the age of 4 years) to 1991 (age of 11 years). Brightness also decreased from 1997 to 2003, as indicated by PC1 scores and surface reflectance of the TM bands 4 (near infrared) and 5 (shortwave infrared). Spectral mixture shade fraction increased from young to old successional stages with strata composition and canopy structure development, whereas NDVI and EVI decreased over time. Because EVI was strongly dependent on near infrared reflectance (r = + 0.96), it was also much more strongly correlated with the shade fraction (r = ?0.93) than NDVI. Except for the image texture mean that decreased from young to old successional stages in TM bands 4 and 5, no clear trend was observed in the remaining texture metrics over the time period of vegetation regeneration. Overall, due to structural-floristic and spectral/textural differences with the PF, the SS site was still distinguishable using Landsat data 30 years after land abandonment. Most of the spectral metric means between PF and SS were significantly different over time at 0.01 significance level, as indicated by ANOVA.     

Predicting forest successional stages using multitemporal Landsat imagery with forest inventory and analysis data

Forest succession is an important ecological process that has profound biophysical, biological and biogeochemical implications in terrestrial ecosystems. Therefore, information on forest successional stages over an extensive forested landscape is crucial for us to understand ecosystem processes, such as carbon assimilation and energy interception. This study explored the potential of using Forest Inventory and Analysis (FIA) plot data to extract forest successional stage information from remotely sensed imagery with three widely used predictive models, linear regression (LR), decision trees (DTs) and neural networks (NNs). The predictive results in this study agree with previous findings that multitemporal Landsat Thematic Mapper (TM) imagery can improve the accuracy of forest successional stage prediction compared to models using a single image. Because of the overlap of spectral signatures of forests in different successional stages, it is difficult to accurately separate forest successional stages into more than three broad age classes (young, mature and old) with reasonable accuracy based on the age information of FIA plots and the spectral data of the plots from Landsat TM imagery. Given the mixed spectral response of forest age classes, new approaches need to be explored to improve the prediction of forest successional stages using FIA data.     

Classifying successional forests using Landsat spectral properties and ecological characteristics in eastern Amazônia

Secondary forests may become increasingly important as temporary reservoirs of genetic diversity, stocks of carbon and nutrients, and moderators of hydrologic cycles in the Amazon Basin as agricultural lands are abandoned and often later cleared again for agriculture. We studied a municipality in northeastern Pará, Brazil, that has been settled for over a century and where numerous cycles of slash and burn agriculture have occurred. The forests were grouped into young (3-6 years), intermediate (10-20 years), advanced (40-70 years), and mature successional stages using 1999 Landsat 7 ETM imagery. Supervised classification of the imagery showed that these forest classes occupied 22%, 13%, 9%, and 6% of the area, respectively. Although this area underwent widespread deforestation many decades ago, forest of some type covers about 50% of the area. Row crops, tree crops, and pastures cover 8%, 20%, and 22%, respectively. The best separation among land covers appeared in a plot of NDVI versus band 5 reflectance. The same groupings of successional forests were derived independently from indices of similarity among tree species composition. Measured distributions of tree height and diameter also covaried with these successional classes, with the young forests having nearly uniform distributions, whereas multiple height and diameter classes were present in the advanced successional forests. Biomass accumulated more slowly in this secondary forest chronosequence than has been reported for other areas, which explains why the 70-year-old forests here were still distinguishable from mature forests using spectral properties. Rates of forest regrowth may vary across regions due to differences in edaphic, climatic, and historical land-use factors, thus rendering most relationships among spectral properties and forest age site-specific. Successional status, as characterized by species composition, biomass, and distributions of heights and diameters, may be superior to stand age as a means of stratifying these forests for characterization of spectral properties.     

Using multiple image endmember spectral mixture analysis to study chaparral regrowth in southern California

This research tested the ability of a multiple endmember (EM) spectral mixture analysis (SMA) approach, applied to multi-temporal Landsat Thematic Mapper (TM) data, to produce realistic and meaningful EM fractions for the study of post-fire regrowth in a southern California chaparral landscape. Eight different image EMs were used, two types for each EM class of interest (green vegetation (GV), non-photosynthetic vegetation (NPV), soil, and shade); the best EM combination was selected for each pixel. These EM fractions were validated with fractions derived from 1?m Airborne Data Acquisition and Registration multi-spectral image data. The EM fractions from the two datasets were similar (r=0.873, 0.776, 0.790 for GV, NPV, and soil, respectively). Chaparral stands were delineated using vegetation type, fire history and slope aspect GIS layers. Mean EM fractions were calculated for each stand, and analysis of variance was performed to determine if EM fractions were different for stands of different age. Short-term trajectories of individual stands appeared to exhibit trends consistent with trends reported in the literature. However, only the youngest and oldest stands were consistently significantly different.     

Assessment of deforestation in the Lower Amazon floodplain using historical Landsat MSS/TM imagery

The floodplain forests bordering the Amazon River have outstanding ecological, economic, and social importance for the region. However, the original distribution of these forests is not well known, since they have suffered severe degradation since the 16th century. The previously published vegetation map of the Amazon River floodplain (Hess et al., 2003), based on data acquired in 1996, shows enormous difference in vegetation cover classes between the regions upstream and downstream of the city of Manaus. The upper floodplain is mostly covered by forests, while the lower floodplain is predominantly occupied by grasses and shrubs.This study assesses deforestation in the Lower Amazon floodplain over a ~ 30 year period by producing and comparing a historical vegetation map based on MSS/Landsat images acquired in the late 1970s with a recent vegetation map produced from TM/Landsat images obtained in 2008. The maps were generated through the following steps: 1) normalization and mosaicking of images for each decade; 2) application of a linear mixing model transformation to produce vegetation, soil and shade fraction-images; and 3) object-oriented image analysis and classification. For both maps, the following classes were mapped: floodplain forest, non-forest floodplain vegetation, bare soil and open water. The two maps were combined using object-level Boolean operations to identify time transitions among the mapped classes, resulting in a map of the land cover change occurred over ~ 30 years. Ground information collected at 168 ground points was used to build confusion matrices and calculate Kappa indices of agreement. A survey strategy combining field observations and interviews allowed the collection of information about both recent and historical land cover for validation purposes. Kappa values (0.77, 0.75 and 0.75) indicated the good quality of the maps, and the error estimates were used to adjust the estimated deforested area to a value of 3457 km² ± 1062 km² (95% CI) of floodplain deforestation over the ~ 30 years.     

Topographic normalization for improving vegetation classification in a mountainous watershed in Northern Thailand

Land cover classifications are adversely affected by shading or topographic effects in mountainous areas in that the spectral properties of an entity in the shade appear to be different from those of the same entity in a sunlit area. Topographic effects can make it especially difficult to distinguish different successional stages of vegetation. The current work uses a simplified topographic normalization method to reduce the topographic effect and to improve land cover classification in a mountainous watershed in northern Thailand. Data used in the study were two Landsat 7 Enhanced Thematic Mapper Plus (ETM+) images acquired on 5 March 2000 and 7 February 2002, a digital elevation model, and Global Positioning Systems (GPS) ground truth data collected in July 2002 consisting of geographic location (latitude/longitude), feature information and ground reference photographs. A supervised land cover classification was conducted on original and normalized images. In general, the classification accuracy of the different successional stages of vegetation was improved in the normalized images.     

Predicting temperate conifer forest successional stage distributions with multitemporal Landsat Thematic Mapper imagery

Forest succession is a fundamental ecological process which can impact the functioning of many terrestrial processes, such as water and nutrient cycling and carbon sequestration. Therefore, knowing the distribution of forest successional stages over a landscape facilitates a greater understanding of terrestrial ecosystems. One way of characterizing forest succession over the landscape is to use satellite imagery to map forest successional stages continuously over a region. In this study we use a forest succession model (ZELIG) and a canopy reflectance model (GORT) to produce spectral trajectories of forest succession from young to old-growth stages, and compared the simulated trajectories with those constructed from Landsat Thematic Mapper (TM) imagery to understand the potential of mapping forest successional stages with remote sensing. The simulated successional trajectories captured the major characteristics of observed regional mean succession trajectory with Landsat TM imagery for Tasseled Cap indices based on age information from the Pacific Northwest Forest Inventory and Analysis Integrated Database produced by Pacific Northwest Research Station, USDA Forest Service. Though the successional trajectories are highly nonlinear in the early years of succession, a linear model fits well the regional mean successional trajectories for brightness and greenness due to significant cross-site variations that masked the nonlinearity over a regional scale (R² = 0.8951 for regional mean brightness with age; R² = 0.9348 for regional mean greenness with age). Regression analysis found that Tasseled Cap brightness and greenness are much better predictors of forest successional stages than wetness index based on the data analyzed in this study. The spectral history based on multitemporal Landsat imagery can be used to effectively identify mature and old-growth stands whose ages do not match with remote sensing signals due to change occurred during the time between ground data collection and image acquisition. Multitemporal Landsat imagery also improves prediction of forest successional stages. However, a linear model on a stand basis has a limited predictive power of forest stand successional stages (adjusted R² = 0.5435 using the Tasseled Cap indices from all four images used in this study) due to significant variations in remote sensing signals for stands at the same successional stage. Therefore, accurate prediction of forest successional stage using remote sensing imagery at stand scale requires accounting for site-specific factors influence remotely sensed signals in the future.     

Coastal wetland vegetation classification with a Landsat Thematic Mapper image

Coastal wetland vegetation classification with remotely sensed data has attracted increased attention but remains a challenge. This paper explored a hybrid approach on a Landsat Thematic Mapper (TM) image for classifying coastal wetland vegetation classes. Linear spectral mixture analysis was used to unmix the TM image into four fraction images, which were used for classifying major land covers with a thresholding technique. The spectral signatures of each land cover were extracted separately and then classified into clusters with the unsupervised classification method. Expert rules were finally used to modify the classified image. This research indicates that the hybrid approach employing sub-pixel information, an analyst's knowledge and characteristics of coastal wetland vegetation distribution shows promise in successfully distinguishing coastal vegetation classes, which are difficult to separate with a maximum likelihood classifier (MLC). The hybrid method provides significantly better classification results than MLC.     

Mapping the land-cover distribution in arid and semiarid urban landscapes with Landsat Thematic Mapper imagery

Mapping the land-cover distribution in arid and semiarid urban landscapes using medium spatial resolution imagery is especially difficult due to the mixed-pixel problem in remotely sensed data and the confusion of spectral signatures among bare soils, sparse density shrub lands, and impervious surface areas (ISAs hereafter). This article explores a hybrid method consisting of linear spectral mixture analysis (LSMA), decision tree classifier, and cluster analysis for mapping land-cover distribution in two arid and semiarid urban landscapes: Urumqi, China, and Phoenix, USA. The Landsat Thematic Mapper (TM) imagery was unmixed into four endmember fraction images (i.e. high-albedo object, low-albedo object, green vegetation (GV), and soil) using the LSMA approach. New variables from these fraction images and TM spectral bands were used to map seven land-cover classes (i.e. forest, shrub, grass, crop, bare soil, ISA, and water) using the decision tree classifier. The cluster analysis was further used to modify the classification results. QuickBird imagery in Urumqi and aerial photographs in Phoenix were used to assess classification accuracy. Overall classification accuracies of 86.0% for Urumqi and 88.7% for Phoenix were obtained, much higher accuracies than those utilizing the traditional maximum likelihood classifier (MLC). This research demonstrates the necessity of using new variables from fraction images to distinguish between ISA and bare soils and between shrub and other vegetation types. It also indicates the different effects of spatial patterns of land-cover composition in arid and semiarid landscapes on urban land-cover classification.     

Land-cover classification in a moist tropical region of Brazil with Landsat Thematic Mapper imagery

This research aims to improve land-cover classification accuracy in a moist tropical region in Brazil by examining the use of different remote-sensing-derived variables and classification algorithms. Different scenarios based on Landsat Thematic Mapper (TM) spectral data and derived vegetation indices and textural images and different classification algorithms, maximum likelihood classification (MLC), artificial neural network (ANN), classification tree analysis (CTA) and object-based classification (OBC), were explored. The results indicate that a combination of vegetation indices as extra bands into Landsat TM multi-spectral bands did not improve the overall classification performance, but the combination of textural images was valuable for improving vegetation classification accuracy. In particular, the combination of both vegetation indices and textural images into TM multi-spectral bands improved the overall classification accuracy (OCA) by 5.6% and the overall kappa coefficient (OKC) by 6.25%. Comparison of the different classification algorithms indicated that CTA and ANN have poor classification performance in this research, but OBC improved primary forest and pasture classification accuracies. This research indicates that use of textural images or use of OBC are especially valuable for improving the vegetation classes such as upland and liana forest classes that have complex stand structures and large patch sizes.     

Detection of vegetation abundance change in the alpine tree line using multitemporal Landsat Thematic Mapper imagery

Detection of alpine tree line change using pixel-based approaches on medium spatial resolution imagery is challenging because of very slow tree sprawl without obvious boundaries. However, vegetation abundance or density in the tree line zones may change over time and such a change may be detected using subpixel-based approaches. In this research, a linear spectral mixture analysis (LSMA)-based approach was used to examine alpine tree line change in the Northern Tianshan Mountains located in Northwestern China. Landsat Thematic Mapper (TM) imagery was unmixed into three fraction images (i.e. green vegetation – GV, shade, and soil) using the LSMA approach. The GV and soil fractions at different years were used to examine vegetation abundance change based on samples in the alpine tree line. The results show that Picea schrenkiana abundance around the top of the forested area increased approximately by 18.6% between 1990 and 2010, but remained stable in the central forest region over this period. Juniperus sabina abundance around the top of the forested area, in the central scrub region, and at the top of the scrub region increased approximately by 19.3%, 8.2%, and 15.6%, respectively. The increased vegetation abundance and decreased soil abundance of both P. schrenkiana and J. sabina indicate vegetation sprawl in the alpine tree line between 1990 and 2010. This research will be valuable for better understanding the impacts of climate change on vegetation change in the alpine tree line of central Asia.     

Land-cover classification in a moist tropical region of Brazil with Landsat TM imagery

This research aims to improve land-cover classification accuracy in a moist tropical region in Brazil by examining the use of different remote sensing-derived variables and classification algorithms. Different scenarios based on Landsat Thematic Mapper (TM) spectral data and derived vegetation indices and textural images, and different classification algorithms - maximum likelihood classification (MLC), artificial neural network (ANN), classification tree analysis (CTA), and object-based classification (OBC), were explored. The results indicated that a combination of vegetation indices as extra bands into Landsat TM multispectral bands did not improve the overall classification performance, but the combination of textural images was valuable for improving vegetation classification accuracy. In particular, the combination of both vegetation indices and textural images into TM multispectral bands improved overall classification accuracy by 5.6% and kappa coefficient by 6.25%. Comparison of the different classification algorithms indicated that CTA and ANN have poor classification performance in this research, but OBC improved primary forest and pasture classification accuracies. This research indicates that use of textural images or use of OBC are especially valuable for improving the vegetation classes such as upland and liana forest classes having complex stand structures and having relatively large patch sizes.     

Mapping forest successional stages following deforestation in Brazilian Amazonia using multi‐temporal Landsat images

Tropical forest successional stages have been mapped previously with multi‐temporal satellite sensor imagery. The precise identification and classification of such stages, however, has proved difficult. This Letter presents a new method for the classification of forest successional stages following deforestation in Brazilian Amazonia. Multi‐temporal Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) and derived fraction images and field data were used in a semi‐automatic classification approach. The results were encouraging and signal the application of the method for the entire Brazilian Amazonia.     

Mapping land cover types in the Amazon Basin using 1 km JERS-1 mosaic

During the Global Rain Forest Mapping (GRFM) project, the JERS-1 SAR (Synthetic Aperture Radar) satellite acquired wall-to-wall image coverage of the humid tropical forests of the world. The rationale for the project was to demonstrate the application of spaceborne L-band radar in tropical forest studies. In particular, the use of orbital radar data for mapping land cover types, estimating the area of floodplains, and monitoring deforestation and forest regeneration were of primary importance. In this paper we examine the information content of the JERS-1 SAR data for mapping land cover types in the Amazon basin. More than 1500 high-resolution (12.5 m pixel spacing) images acquired during the low flood period of the Amazon river were resampled to 100 m resolution and mosaicked into a seamless image of about 8 million km², including the entire Amazon basin. This image was used in a classifier to generate a 1 km resolution land cover map. The inputs to the classifier were 1 km resolution mean backscatter and seven first-order texture measures derived from the 100 m data by using a 10 x 10 independent sampling window. The classification approach included two interdependent stages. First, a supervised maximum a posteriori Baysian approach classified the mean backscatter image into five general cover categories: terra firme forest (including secondary forest), savanna, inundated vegetation, open deforested areas and open water. A hierarchical decision rule based on texture measures was then applied to attempt further discrimination of known subcategories of vegetation types based on taxonomic information and woody biomass levels. True distributions of the general categories were identified from the RADAMBRASIL project vegetation maps and several field studies. Training and validation test sites were chosen from the JERS-1 image by consulting the RADAM vegetation maps. After several iterations and combining land cover types, 14 vegetation classes were successfully separated at the 1 km scale. The accuracy of the classification methodology was estimated to be 78% when using the validation sites. The results were also verified by comparison with the RADAM- and AVHRR-based 1 km resolution land cover maps.     

Mapping montane tropical forest successional stage and land use with multi-date Landsat imagery

When mapping land cover with satellite imagery in montane tropical regions, varying illumination angles and ecological zones can obscure the differences between spectral responses of old-growth forest, secondary forest and agricultural lands. We used multi-date, Landsat Thematic Mapper (TM) imagery to map secondary forests, agricultural lands and old-growth forests in the Talamanca Mountain Range in southern Costa Rica. With stratification by illumination and ecological zone, the overall accuracy for this classification was 87% with a Kappa coefficient of 0.83. We also examined spectral responses to forest successional stage, ecological zone and aspect illumination for the TM Tasselled Cap indices, TM (2 x 6)/7, TM 4/5 and TM difference bands, and whether using digital data from multiple decades improved classification accuracy. Digital maps of ecological zones should be useful for large-scale mapping of land use and forest successional stage in complex montane regions such as those in Central America.     

Remote sensing of industrial impact on Arctic vegetation around Noril'sk,northern Siberia: Preliminary results

This study applies remote sensing techniques for monitoring non-ferrous metal smelting impacts in the extreme environment of northern Siberia. Ground and at-satellite reflectance and normalized difference vegetation index (NDVI) values for different vegetation types have been compared and a hybrid supervised-unsupervised classification of Landsat TM data performed, based on field and ancillary data. This has allowed us to distinguish several degrees of vegetation damage in tundra and forests. However, it was difficult to differentiate between some significant classes, such as damaged grass tundra and sparse dead larch forests with a grass understorey. We suggest possible refinement of our results, including the combination of images taken at different phenological stages and from different sensors. However, it should be noted that the north-Siberian environment presents unusually severe limitations of optical-infrared satellite observation possibilities and problems in imagery interpretation. Standard indicators of vegetation vigour, such as NDVI, widely applied at lower latitudes, become less informative in highly variable tundra and pre-tundra ecosystems.     

Mapping the regional extent of tropical forest regeneration stages in the Brazilian Legal Amazon using NOAA AVHRR data

The study aimed to map several stages of tropical forest regeneration across the Brazilian Legal Amazon using 1.1 km NOAA AVHRR data. Regenerating forest extent was defined using an unsupervised classification of AVHRR channels 1, 2 and 3 and the Global Environment Monitoring Index (GEMI). A method for discriminating four forest regeneration stages was then developed, based on relationships between AVHRR channels 1, 2 and 3 and forest age. This method was applied to AVHRR data to map forests associated with Stages I (early colonization phase, open canopy, < 5 years), II (closed, singlelayered canopy, 5-9 years), III (closed canopy with structural development, 9-20 years) and IV (closed multilayered canopy, > 20 years). The maps provided new regional estimates of regenerating forest for the Legal Amazon and indicated that, over the period 1991 to 1994, approximately 35.8% (157 973 km2) of the total deforested area of 440 186 km2 (estimated for 1992) supported regenerating forest, with 48% of these forests aged at less than 5 years. The study concluded that AVHRR data has an important role in mapping and monitoring tropical forest regeneration. The datasets generated provide valuable input to models of regional carbon flux. For example, Grace et al . (1995a, b) reported net annual CO2 absorption 8.5 2.0 moles m 2 for mature forests in south-west Amazonia suggesting     

Relationships among soil fertility dynamics and remotely sensed measures across pasture chronosequences in Rondônia, Brazil

This study analyzed the relationships between soil fertility and remotely sensed measures over three pasture chronosequence sites in the state of Rondônia, in the western Brazilian Amazon region. Remotely sensed measures included shade, nonphotosynthetic vegetation (NPV), green vegetation (GV) and soil (derived from spectral mixture analysis), and the normalized difference vegetation index (NDVI). These were correlated against soil fertility parameters such as phosphorus, potassium, calcium, and base saturation. In temporal analysis, it was observed that NPV dominated the spectral responses of pasture canopies and tended to increase with pasture age as well. The increase of NPV appeared to be related to the decline of soil fertility, but soil texture variation also played a role. In the correlation analysis, soil P, known as the most limiting nutrient for pasture productivity, showed the highest correlation with remotely sensed measures, followed by soil K and base saturation. However, this result was not observed at the sites where nutrient availability was very low.     

Characterization of pasture biophysical properties and the impact of grazing intensity using remotely sensed data

Remote sensing has the potential of improving our ability to map and monitor pasture degradation. Pasture degradation is one of the most important problems in the Amazon, yet the manner in which grazing intensity, edaphic conditions and land‐use age impact pasture biophysical properties, and our ability to monitor them using remote sensing is poorly known. We evaluate the connection between field grass biophysical measures and remote sensing, and investigate the impact of grazing intensity on pasture biophysical measures in Rondônia, in the Brazilian Amazon. Above ground biomass, canopy water content and height were measured in different pasture sites during the dry season. Using Landsat Thematic Mapper (TM) data, four spectral vegetation indices and fractions derived from spectral mixture analysis, i.e., Non‐Photosynthetic Vegetation (NPV), Green Vegetation (GV), Soil, Shade, and NPV + Soil, were calculated and compared to field grass measures. For grazed pastures under dry conditions, the Normalized Difference Infrared Index (NDII5 and NDII7), had higher correlations with the biophysical measures than the Normalized Difference Vegetation Index (NDVI) and the Soil‐Adjusted Vegetation Index (SAVI). NPV had the highest correlations with all field measures, suggesting this fraction is a good indicator of pasture characteristics in Rondônia. Pasture height was correlated to the Shade fraction. A conceptual model was built for pasture biophysical change using three fractions, i.e., NPV, Shade and GV to characterize possible pasture degradation processes in Rondônia. Based upon field measures, grazing intensity had the most significant impact on pasture biophysical properties compared to soil order and land‐use age. The impact of grazing on pastures in the dry season could be potentially measured by using remotely sensed measures such as NPV.     

Towards a comprehensive evaluation of V-I-S sub-pixel fractions and land surface temperature for urban land-use classification in the USA

Remote-sensing image classification based on the vegetation–impervious surface–soil (V-I-S) model and land-surface temperature (LST) has proved to be more efficient in characterizing the urban landscape than conventional spectral-based classification. However, current literature emphasizes discussion of the classifier's accuracy improvement achieved by the input of V-I-S fractions and LST over conventional spectral-based classification while ignoring the stability evaluation. Hence, this study proposes an evaluation framework for exploring the superiority of the input features and the stability of classifiers by integrating statistical randomization techniques and a kappa-error diagram. The evaluation framework was applied to case studies for demonstrating the superiority of V-I-S fractions and LST in the context of urban land-use classification with five different types of classifiers, including the maximum likelihood classifier (MLC), the tree classifier, the Bagging classifier, the random forest (RF) and the support vector machine (SVM). It followed that the use of V-I-S fractions and LST (1) could alleviate the ‘salt and pepper’ effect; (2) is preferred by tree and tree-based ensembles for branch splitting; (3) could produce classification trees with less complexity; (4) could benefit the stability of classifiers in addition to the accuracy improvement; and (5) could allow histograms following nearly normal distribution in its feature space, which boosts the performance of MLC. It is shown that MLC becomes comparable with modern classifiers when trained with V-I-S fractions and LST combination. Because of its adequacy and simplicity, MLC is recommended for urban land-use classification when V-I-S fractions and LST are used as the only input features. However, replacing them with, or including, the band reflectance might degrade MLC. A direct use of spectral band reflectance is not recommended for any of the classification approaches being considered in this study, except for SVM, which is the most robust classifier as it has a consistently high performance for all the input feature combinations. We recommend using tree-based ensemble classifiers or SVM when V-I-S fractions and LST as well as the band reflectance are all used in the classification. The proposed evaluation framework can also be applied to the assessment of input features and classifiers in other remote-sensing classification endeavours.