The suitability of multi-temporal web-enabled Landsat data NDVI for phenological monitoring – a comparison with flux tower and MODIS NDVI

Three years of flux-tower-derived normalized difference vegetation index (NDVI) data were compared with contemporaneous 30 m web-enabled Landsat data (WELD) and with 500 m Moderate-Resolution Imaging Spectroradiometer (MODIS) nadir bidirectional reflectance distribution function-adjusted reflectance (NBAR) NDVI data to assess the relative suitability of these different resolutions of freely available satellite data for phenological monitoring. Comparisons were made at two flux tower sites in the United States with average to above average cloud cover. The WELD 30 m NDVI data were found to have higher correlation with the flux tower NDVI data than the MODIS 500 m NBAR NDVI data. The dates of vegetation green-up onset and maximum-greenness onset, derived using an established phenological metric extraction methodology, were generally closer between the flux tower and WELD NDVI data than between the flux tower and MODIS NBAR data. These results indicate that the WELD NDVI time series is suitable for 30 m scale phenological monitoring.     

A semi-automated approach for the generation of a new land use and land cover product for Germany based on Landsat time-series and Lucas in-situ data

Information on land cover and land use at high spatial resolutions is essential for advancing earth system science as well as for environmental monitoring to support decision-making and reporting processes. In view of this, we present the first version of the DFD Land Use and Land Cover Product for Germany, DFD-LULC_DE, for the year 2014, generated from 702 Landsat-7 and Landat-8 scenes at 30 m resolution. The results were derived based on a fully automated preprocessing chain that integrates data acquisition, radiometric, atmospheric and topographic correction, as well as spectral–temporal feature extraction for all Landsat surface reflectance bands, brightness temperature and various spectral indices. The classification followed a two-step approach: first, an initial classification is performed using a Random Forest classifier trained on ground truth data obtained from the LUCAS survey of EUROSTAT, followed by a semi-automated sampling of additional training data to further improve the initial classification results. Automatic selection of appropriate training samples is based on the vote entropy derived from the initial classification, thereby keeping manual user interaction low. The approach demonstrated is promising, also with respect to a European wide application, and contributes towards the advancement and enhancement of the DLR-DFD’s processing chains, which are directed towards the generation of land cover products at regular intervals being of central importance to related land monitoring and reporting services.     

A new approach for mapping regional land cover and the application of this approach in Australia

Detailed and accurate land cover data are widely used for various purposes, such as global land use change detection. This study aimed to investigate Australian land use and cover change by using 774 Landsat scenes in 2000 and 2010. The reference data included pictures or high-resolution images from Google Earth, global land cover data, Shuttle Radar Topography Mission data, and other literature. Australian land cover was automatically classified into nine main classes, namely cropland, forest, grassland, shrubs, wetland, water, artificial covering, bare land, and others, by using a global land cover decision tree developed for computer-based image classification. Obvious mistakes were subsequently modified manually. Finally, the classifications were found to have overall accuracies of 93.24% and 92.79% evaluated by using 2820 and 3820 sample points, respectively.     

Quantifying canopy fractional cover and change in Fanjingshan National Nature Reserve,China using multi-temporal Landsat imagery

This letter explores factors affecting the quantification and mapping of forest canopy fractional cover (CFC), and explores causes of CFC change. CFC was quantified using a simple linear mixture model based on the modified soil-adjusted vegetation index (MSAVI) derived from Landsat TM surface reflectance data of Fanjingshan National Nature Reserve (FNNR) in China. Different soil and vegetation endmembers were tested to analyse the sensitivity of the mixture model. Illumination effects due to topographical variability are found to influence MSAVI and therefore CFC estimates. Implementing an illumination stratification that selects different closed canopy endmembers for different topographic-related illumination strata generally minimizes these effects. The spatial distribution and possible causes of CFC change were examined. Most changes in CFC over the 15-year study period appear to have resulted from anthropogenic activities, at least based on the precision constraints of Landsat-derived CFC change estimates and limited high spatial resolution imagery used in a mostly visual verification of patches with low CFC and reduction in CFC between image dates.     

Testing the local applicability of MODIS BRDF parameters for correcting Landsat TM imagery

Landsat Thematic Mapper (TM) imagery is widely used for large scale multi-temporal monitoring of the land surface. One source of variability in this imagery, which is not related to variation in the land surface, is the result of the bi-directional reflectance distribution function (BRDF). This causes variation in the measured reflectance as a function of the angular configuration of the sun and sensor. To remove this variation from the imagery, a correction can be applied for BRDF. Some authors have proposed the use of the BRDF parameterization derived for the Moderate Resolution Imaging Spectroradiometer (MODIS; NASA, Washington, DC, USA) instrument as a means of correction for BRDF in Landsat TM imagery. However, since the spatial scale of the two instruments is substantially different, it is not clear up to what extent the MODIS BRDF parameters are applicable to Landsat TM. This study provides a direct empirical test of this, over the eastern Australian landscape, and tests whether the MODIS BRDF parameters are applicable to characterize BRDF at a local scale of a Landsat TM pixel, or simply at a global scale. The results suggest that the MODIS BRDF parameters effectively characterize the BRDF of the Landsat imagery at a global average scale, but, at least in the Australian context, do not capture any further information about the BRDF of more local regions, or individual Landsat pixels.     

An automated method for the detection of emperor penguin colonies from Landsat 8 imagery

Detailed information on the emperor penguin colonies is crucial for estimating total populations and analysing population migration. This study presents a new method for detecting colonies of emperor penguins by identifying areas covered with their faeces using Landsat 8 data. Top of Atmosphere (TOA) reflectance and Brightness Temperature (BT) of Landsat images are used as inputs. This method first uses normalized spectral indexes (normalized difference water index (NDWI), normalized difference faeces index (NDFI), normalized difference snow index (NDSI)), band ratios and an individual band reflectance to produce probability masks for areas covered by faeces differentiating from other land cover types. Then, after eliminating those pixels that have abnormal elevations and performing a median filter, the probability masks for those areas covered by faeces are used to derive the corresponding polygons. Subsequently, the geometric centre of the polygons of those areas covered by faeces is used as the location for a corresponding colony. For a widely distributed set of data around the Ross and Somov Seas, the overall classification accuracy is as high as 91% with a small standard deviation of 0.12.     

Mapping karst rocky desertification using Landsat 8 images

Karst rocky desertification (KRD) is a serious ecological problem in southwest China. Various remote sensing techniques are available for investigating KRD. Landsat 8 Operational Land Imager (OLI) data, acquired in September (summer) 2013 and January (winter) 2014, were used to analyse the role of vegetation fraction, bedrock exposure, and slope on KRD classification. Then, the decision tree and fuzzy maximum likelihood methods were compared, using the above-mentioned factors, to verify the potential of Landsat 8 OLI data in monitoring KRD. The results show that these factors correlate well with the degree of KRD and that the addition of these factors into the classifier improved accuracy from 84.23% to 91.71%. Thus, Landsat 8 OLI data can be adapted for the monitoring of KRD, which will be useful for the 2015 Third National Desertification Survey.     

Discrete classification approach to land cover and land use change identification based on Landsat image time sequences

Dense multi-temporal stacks of Landsat imagery have most commonly been exploited to identify land cover and land use changes (LCLUC) based on detection of abrupt changes in continuous value spectral indices. In this study, a discrete classification approach to LCLUC identification based on stable training sites is tested on a nine-date, 4-year Landsat-7 ETM + time sequence for a study area in Ghana that is prone to cloud cover. Change to Built cover, as an indication of urban expansion, was identified for over 70% of testing units when a spatial-temporal majority filter that ignored No Data values from clouds, cloud shadows and sensor effects was applied. More important, relatively stable LCLU maps were generated and No Data effects should not limit the potential of the approach for longer-term retrospective analyses or monitoring of LCLUC in cloud-prone regions.     

Comparing bright-target surface spectral-reflectance estimates obtained from IRS P6 LISS III to those obtained from Landsat 5 TM and Landsat 7 ETM+

Since 2008 there have been a limited number of Landsat 5 thematic mapper (TM) images acquired between April and October in Australia. Consequently, TM imagery may not be available at the desired time of year for some monitoring applications. IRS (Indian Remote Sensing) P6 LISS (Linear Imaging and Self Scanner) III imagery has been acquired over Australia since 2008 and represents an alternative, Landsat-like, data source to fill the Landsat 5 TM temporal gap. To be useful for the continuation of long-term monitoring, the LISS III imagery needs to provide similar surface-reflectance estimates to Landsat 5 TM. A time series of spatially averaged sensor-radiance estimates for 2008 was derived from Landsat 5 TM, Landsat 7 enhanced thematic mapper plus (ETM+) and IRS P6 LISS III imagery for two highly reflective, spectrally invariant, claypans in Queensland, Australia. The radiance values were converted to surface reflectance using the atmospheric transfer modelling code 6S. Adjustment factors, to account for the spectral band difference effects between sensors, were computed from field-measured reflectance spectra. The LISS III surface-reflectance estimates were found to be consistently lower than the Landsat estimates. The difference between the IRS P6 LISS III reflectances and the median Landsat 5 TM reflectances were approximately 20%, 22%, 12% and 3.5% for Landsat bands 2, 3, 4 and 5, respectively. Further research is required to determine whether updated calibration parameters for the LISS III sensor are required.     

Towards an operational method for land surface temperature retrieval from Landsat 8 data

Land surface temperature (LST) is a key parameter in the physics of land surfaces through the processes of energy and water exchange with the atmosphere. For Landsat data with only one thermal infrared channel (Landsat 4 to Landsat 7), LST cannot actually be retrieved, and external data sources, such as meteorological observations or Moderate Resolution Imaging Spectroradiometer (MODIS) data, are needed to obtain the water vapour content parameter (an important input parameter for the LST retrieval algorithm); this results in limitations on deriving LST. However, the band designations of the Landsat 8 sensors enable the derivation of LST from the Landsat 8 data. This article demonstrates an LST retrieval methodology that makes use of only Landsat 8 image data. In this methodology, the split-window covariance-variance ratio (SWCVR) technique is introduced to derive water vapour content from Landsat 8. A comparison between the retrieved LST and the in situ LST measurements shows good accuracy, with a root mean squared error (RMSE) of 0.83 K. The fact that the proposed LST estimation method utilizing solely Landsat 8 image data does not rely on any external data is a significant advantage for the development of an operational Landsat 8 LST product generating system.     

Spectral-radiometric differentiation of non-photosynthetic vegetation and soil within Landsat and Sentinel 2 wavebands

We explore the comprehensive differentiation of non-photosynthetic vegetation (NPV) and soil using Landsat and Sentinel 2 wavebands through a spectral library approach. NPV and soil spectra from online spectral libraries and an Airborne Visible Infrared Imaging Spectrometer (AVIRIS) scene were convolved to Landsat 5 Thematic Mapper (TM), Landsat 8 Operational Land Imager (OLI), and Sentinel 2 bands. Several spectral-radiometric measurements, including the spectral reflectance of analogous Landsat and Sentinel bands, and a suite of spectral indices were tested for the separation of NPV and soil. Reflectance of individual bands is similar between the two categories, and vegetation indices such as Normalized Difference Vegetation Index (NDVI) and the Soil Adjusted Vegetation Index (SAVI) are incapable of such differentiation. For both reference and image spectra, the normalized difference tillage index (NDTI) based on the two shortwave infrared bands of Landsat and Sentinel instruments performed best among all spectral measurements. Classification results suggest an NDTI value of 0.1 to be a general threshold for separating NPV and soil, with higher values associated with NPV. Further tests based on AVIRIS-convolved imagery show that NDTI can dichotomize NPV and soil if either fractional cover is no less than 50%.     

Automated detection of impervious surfaces using night-time light and Landsat images based on an iterative classification framework

ABSTRACT

A novel workflow for automated detecting of impervious surface by using night-time light and Landsat images at the individual city scale is proposed. This approach is composed by of three steps. In the beginning, urban, peri-urban and rural regions are detected from the night-time light image by a contour line algorithm. Then, using Landsat TM image, region-specific spectral index analysis is employed to generate initial training samples of urban land covers. Finally, an iterative classification framework is applied to select new training samples by integrating spectral and spatial information and to obtain the final mapping result. Experimental results of two cities show that the proposed method produces higher classification accuracy than the ones using the manual-sampling methods. Moreover, further validations suggested that the spatial information is able to effectively increase the producer’s accuracy of impervious surface. This automated approach is potentially important for large-scale regional impervious surface mapping and application.     

A new index for mapping built-up and bare land areas from Landsat-8 OLI data

Remote sensing is a useful technology for monitoring the spatial distribution and expansion of built-up and bare land areas. One effective approach, known as the normalized difference built-up index (NDBI), has been promoted for identifying built-up areas based on Landsat Thematic Mapper (TM)/Enhanced Thematic Mapper (ETM+) data. The successful launch of the Landsat-8 satellite has made possible the continued acquisition of high-quality data that meet requirements for observing land use and land cover (LULC), and whether or not NDBI approach can be used with Landsat-8 data needs further verification. In this study, we researched the reflectance spectral characteristics of different land cover types in different bands of Landsat-8 operational land imager (OLI) data and found that the trend of some built-up areas in the OLI data from the near-infrared band to the shortwave infrared band is different from that in the TM/ETM+ data. This different trend made the conventional NDBI approach unsuitable for Landsat-8 OLI data. We propose a new index, called the build-up and bare land areas index, for transforming Landsat-8 OLI data to map built-up and bare land areas automatically. This new index was used to detect the built-up and bare land areas in Zhengzhou (Henan, China). The accuracy assessment indicates that our index has much higher accuracy (90.8%) than the conventional NDBI approach does (57.4%).     

Framework for monitoring the conversion of cultivated land to construction land using SAR image time series

The land cover change from cultivated land to construction land is a world issue since the urbanization process is extensively studied around the world. Chengdu, China, is a representative urbanization area, where cloud cover is very high most of the time, restricting the use of visible and near-infrared satellite data. Here, we present a novel framework for land change monitoring based on synthetic aperture radar (SAR) time series, which comprises three key components: (1) construction of pixel-level SAR image time series; (2) spatio-temporal similarity analysis based on morphological-structural characteristics of time series; and (3) iterative binary partition mean square error model analysis to ascertain change nodes. Experimental results showed that the proposed framework could effectively extract the change nodes and change pixels, with correctness of 85.82% and completeness of 84.78%, outperforming the time-series-only (non-spatial) method, as well as traditional classification methods, and the same framework using shorter Landsat TM image time series.     

A comparison of Landsat ETM+ and high-resolution aerial orthophotos to map urban/suburban forest cover in Massachusetts,USA

This article examines the extent to which L(ow)-spatial resolution Landsat Enhanced Thematic Mapper Plus (ETM+) imagery can be used to map urban/suburban forest cover in comparison with H(igh)-spatial resolution (less than 1 m) digital aerial orthophotos from the same study area and time period. This research has practical implications for resource managers, government agencies and forestry researchers interested in mapping large-area urban/suburban forests because Landsat imagery is more accessible, has an extensive historical archive, has broader spatial and temporal coverage and is more cost efficient than H-resolution aerial orthophotos. Classification tree results suggest that Landsat ETM+?imagery is adequate for mapping larger, contiguous patches of forest (i.e. small forest patches greater than 2 acres) in urban/suburban settings, but its spatial resolution is too coarse to accurately map spatially complex residential areas in urban/suburban landscapes.     

Development and validation of the global surface type data product from S-NPP VIIRS

Accurate representation of actual terrestrial surface types at regional to global scales is an important element for many applications. Based on National Aeronautics and Space Administration Moderate Resolution Imaging Spectroradiometer land cover algorithms, a global surface-type product from observations of the Visible Infrared Imaging Radiometer Suite (VIIRS) on board the Suomi National Polar-orbiting Partnership, provides consistent global land cover classification map for various studies, such as land surface modelling for numerical weather predictions, land management, biodiversity and hydrological modelling, and carbon and ecosystem studies. This letter introduces the development and validation of the VIIRS global surface-type product using the land cover classification scheme of the International Geosphere-Biosphere Programme. Surface reflectance data from VIIRS were composited into monthly data and then into annual metrics. The C5.0 decision tree classifier was used to determine the surface type for each pixel in a 1 km grid. To quantitatively evaluate accuracies of the new surface type product, a visual interpretation-based validation was performed in which high-resolution satellite images and other ancillary data were used as the reference. The validation result based on the large validation data set indicated that (78.64 ± 0.57)% overall classification accuracy was achieved.     

Super-resolution land cover mapping by deep learning

Super-resolution mapping (SRM) is a technique to estimate a fine spatial resolution land cover map from coarse spatial resolution fractional proportion images. SRM is often based explicitly on the use of a spatial pattern model that represents the land cover mosaic at the fine spatial resolution. Recently developed deep learning methods have considerable potential as an alternative approach for SRM, based on learning the spatial pattern of land cover from existing fine resolution data such as land cover maps. This letter proposes a deep learning-based SRM algorithm (DeepSRM). A deep convolutional neural network was first trained to estimate a fine resolution indicator image for each class from the coarse resolution fractional image, and all indicator maps were then combined to create the final fine resolution land cover map based on the maximal value strategy. The results of an experiment undertaken with simulated images show that DeepSRM was superior to conventional hard classification and a suite of popular SRM algorithms, yielding the most accurate land cover representation. Consequently, methods such as DeepSRM may help exploit the potential of remote sensing as a source of accurate land cover information.     

Relationship between canopy height and Landsat ETM+ response in lowland Amazonian rainforest

This letter investigates the influence of within-pixel variation of canopy height on the spectral response recorded in Landsat Enhanced Thematic Mapper (ETM+) data for tropical rainforest. Forest canopy height is derived from airborne, small-footprint LiDAR data acquired using a Leica ALS50 II system. The field site is in the Tambopata National Reserve, in Peruvian Amazonia, where forest types include regenerating, swamp, floodplain and terra firme. For individual Landsat ETM+?bands, the strongest correlation for maximum, mean and standard deviation of canopy height occurred with ETM+?Band 4 (near infrared) for regenerating, floodplain and terra firme forest, and with ETM+?Band 5 (middle infrared) for swamp forest. For normalized difference band indices, ND42 and ND43 (i.e. the Normalized Difference Vegetation Index, NDVI) showed strong correlation with both mean and maximum canopy height for regenerating and terra firme forest, and with maximum and standard deviation of canopy height for floodplain forest. The palm-dominated swamp forest showed weaker relationships, with the strongest occurring for ND45 and ND52 with mean canopy height. Many papers have identified middle-infrared bands as being most sensitive to tropical rainforest structure, although these have often focussed on young regenerative forests. By focussing on older regenerative forest (of >25 years since land abandonment) and mature rainforest types, this work has shown that there is considerable variation with how structure may influence spectral reflectance and lends support to the hypothesis that canopy height distribution and shadowing effects caused by canopy complexity and the presence of emergent trees is what most significantly influences spectral response for tropical rainforests.     

High-resolution landcover classification using Random Forest

Potential data sets for landcover classification, such as Landsat (or pre-processed data such as the National Land Cover Dataset (NLCD)), are often too coarse for fine-scale research needs or are cost-prohibitive (Quickbird, Ikonos and Geoeye). Repeated attempts at classifying high spatial resolution data, National Agricultural Imagery Program (NAIP) imagery, based on traditional techniques, such as a maximum likelihood supervised classification, have failed to produce a product with sufficient accuracy. We used the ensemble Random Forests (RFs) classifier to classify landcover at 1 m resolution using 2009 NAIP imagery in south-eastern Wyoming. We classified riparian areas within a 225 km² study area, at 1 m spatial resolution, using RFs with emphasis on riparian corridors that yielded a land cover map with overall accuracy of 81% and a kappa coefficient of 79%. Users’ accuracy of important riparian vegetation species, aspen, riparian grasses and willow were 79%, 81% and 83%, respectively. Techniques presented in this paper, which exploit free NAIP imagery for landcover classification, represent an inexpensive and reliable alternative to purchasing commercial imagery when high spatial resolution landcover data is needed.     

Limitations of the dense dark vegetation method for aerosol retrieval under Australian conditions

The use of dense dark vegetation (DDV) for atmospheric aerosol correction of Landsat imagery is investigated for Australian conditions. Aerosol optical depth (AOD) measurements from sun photometers are used as a reference data set and compared against estimates of AOD derived from Landsat imagery using the DDV method. The DDV method makes assumptions that the vegetation is sufficiently dark and the ratio between bottom-of-atmosphere reflectances at different wavelengths is constant. These assumptions were tested using Landsat-5 Thematic Mapper (TM) imagery corrected with AOD measured by field-based sun photometers on the AErosol RObotic NETwork (AERONET) network. The assumptions were found to be correct only for one of the three locations studied. In other locations, the spatial and temporal variability of the vegetation and its relative brightness makes the method unsuitable.