Forest cover classification using Landsat ETM+ data and time series MODIS NDVI data

Forest cover plays a key role in climate change by influencing the carbon stocks, the hydrological cycle and the energy balance. Forest cover information can be determined from fine-resolution data, such as Landsat Enhanced Thematic Mapper Plus (ETM+). However, forest cover classification with fine-resolution data usually uses only one temporal data because successive data acquirement is difficult. It may achieve mis-classification result without involving vegetation growth information, because different vegetation types may have the similar spectral features in the fine-resolution data. To overcome these issues, a forest cover classification method using Landsat ETM+ data appending with time series Moderate-resolution Imaging Spectroradiometer (MODIS) Normalized Difference Vegetation Index (NDVI) data was proposed. The objective was to investigate the potential of temporal features extracted from coarse-resolution time series vegetation index data on improving the forest cover classification accuracy using fine-resolution remote sensing data. This method firstly fused Landsat ETM+ NDVI and MODIS NDVI data to obtain time series fine-resolution NDVI data, and then the temporal features were extracted from the fused NDVI data. Finally, temporal features combined with Landsat ETM+ spectral data was used to improve forest cover classification accuracy using supervised classifier. The study in North China region confirmed that time series NDVI features had significant effects on improving forest cover classification accuracy of fine resolution remote sensing data. The NDVI features extracted from time series fused NDVI data could improve the overall classification accuracy approximately 5% from 88.99% to 93.88% compared to only using single Landsat ETM+ data.     

Global, 30-m resolution continuous fields of tree cover: Landsat-based rescaling of MODIS vegetation continuous fields with lidar-based estimates of error

Abstract

We developed a global, 30-m resolution dataset of percent tree cover by rescaling the 250-m MOderate-resolution Imaging Spectroradiometer (MODIS) Vegetation Continuous Fields (VCF) Tree Cover layer using circa- 2000 and 2005 Landsat images, incorporating the MODIS Cropland Layer to improve accuracy in agricultural areas. Resulting Landsat-based estimates maintained consistency with the MODIS VCF in both epochs (RMSE =8.6% in 2000 and 11.9% in 2005), but showed improved accuracy in agricultural areas and increased discrimination of small forest patches. Against lidar measurements, the Landsat-based estimates exhibited accuracy slightly less than that of the MODIS VCF (RMSE=16.8% for MODIS-based vs. 17.4% for Landsat-based estimates), but RMSE of Landsat estimates was 3.3 percentage points lower than that of the MODIS data in an agricultural region. The Landsat data retained the saturation artifact of the MODIS VCF at greater than or equal to 80% tree cover but showed greater potential for removal of errors through calibration to lidar, with post-calibration RMSE of 9.4% compared to 13.5% in MODIS estimates. Provided for free download at the Global Land Cover Facility (GLCF) website (www.landcover.org), the 30-m resolution GLCF tree cover dataset is the highest-resolution multi-temporal depiction of Earth's tree cover available to the Earth science community.     

Updating Landsat-based forest cover maps with MODIS images using multiscale spectral-spatial-temporal superresolution mapping

With the high deforestation rates of global forest covers during the past decades, there is an ever-increasing need to monitor forest covers at both fine spatial and temporal resolutions. Moderate Resolution Imaging Spectroradiometer (MODIS) and Landsat series images have been used commonly for satellite-derived forest cover mapping. However, the spatial resolution of MODIS images and the temporal resolution of Landsat images are too coarse to observe forest cover at both fine spatial and temporal resolutions. In this paper, a novel multiscale spectral-spatial-temporal superresolution mapping (MSSTSRM) approach is proposed to update Landsat-based forest maps by integrating current MODIS images with the previous forest maps generated from Landsat image. Both the 240 m MODIS bands and 480 m MODIS bands were used as inputs of the spectral energy function of the MSSTSRM model. The principle of maximal spatial dependence was used as the spatial energy function to make the updated forest map spatially smooth. The temporal energy function was based on a multiscale spatial-temporal dependence model, and considers the land cover changes between the previous and current time. The novel MSSTSRM model was able to update Landsat-based forest maps more accurately, in terms of both visual and quantitative evaluation, than traditional pixel-based classification and the latest sub-pixel based super-resolution mapping methods The results demonstrate the great efficiency and potential of MSSTSRM for updating fine temporal resolution Landsat-based forest maps using MODIS images.     

Monitoring aboveground forest biomass dynamics over three decades using Landsat time-series and single-date inventory data

Understanding forest biomass dynamics is crucial for carbon and environmental monitoring, especially in the context of climate change. In this study, we propose a robust approach for monitoring aboveground forest biomass (AGB) dynamics by combining Landsat time-series with single-date inventory data. We developed a Random Forest (RF) based kNN model to produce annual maps of AGB from 1988 to 2017 over 7.2 million ha of forests in Victoria, Australia. The model was internally evaluated using a bootstrapping technique. Predictions of AGB and its change were then independently evaluated using multi-temporal Lidar data (2008 and 2016). To understand how natural and anthropogenic processes impact forest AGB, we analysed trends in relation to the history of disturbance and recovery. Specifically, change metrics (e.g., AGB loss and gain, Years to Recovery - Y2R) were calculated at the pixel level to characterise the patterns of AGB change resulting from forest dynamics. The imputation model achieved a RMSE value of 132.9 Mg ha⁻¹ (RMSE% = 46.3%) and R² value of 0.56. Independent assessments of prediction maps in 2008 and 2016 using Lidar-based AGB data achieved relatively high accuracies, with a RMSE of 108.6 Mg ha⁻¹ and 135.9 Mg ha⁻¹ for 2008 and 2016, respectively. Annual validations of AGB maps using un-changed, homogenous Lidar plots suggest that our model is transferable through time (RMSE ranging from 109.65 Mg ha⁻¹ to 112.27 Mg ha⁻¹ and RMSE% ranging from 25.38% to 25.99%). In addition, changes in AGB values associated with forest disturbance and recovery (decrease and increase, respectively) were captured by predicted maps. AGB change metrics indicate that AGB loss and Y2R varied across bioregions and were highly dependent on levels of disturbance severity (i.e., a greater loss and longer recovery time were associated with a higher severity disturbance). On average, high severity fire burnt from 200 Mg ha⁻¹ to 550 Mg ha⁻¹ of AGB and required up to 15 years to recover while clear-fell logging caused a reduction in 250 Mg ha⁻¹ to 600 Mg ha⁻¹ of AGB and required nearly 20 years to recover. In addition, AGB within un-disturbed forests showed statistically significant but monotonic trends, suggesting a mild gradual drop over time across most bioregions. Our methods are designed to support forest managers and researchers in developing forest monitoring systems, especially in developing regions, where only a single date forestry inventory exists.     

Evaluating land cover changes in Eastern and Southern Africa from 2000 to 2010 using validated Landsat and MODIS data

In this study, we assessed land cover land use (LCLU) changes and their potential environmental drivers (i.e., precipitation, temperature) in five countries in Eastern & Southern (E&S) Africa (Rwanda, Botswana, Tanzania, Malawi and Namibia) between 2000 and 2010. Landsat-derived LCLU products developed by the Regional Centre for Mapping of Resources for Development (RCMRD) through the SERVIR (Spanish for “to serve”) program, a joint initiative of NASA and USAID, and NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS) data were used to evaluate and quantify the LCLU changes in these five countries. Given that the original development of the MODIS land cover type standard products included limited training sites in Africa, we performed a two-level verification/validation of the MODIS land cover product in these five countries. Precipitation data from CHIRPS dataset were used to evaluate and quantify the precipitation changes in these countries and see if it was a significant driver behind some of these LCLU changes. MODIS Land Surface Temperature (LST) data were also used to see if temperature was a main driver too.Our validation analysis revealed that the overall accuracies of the regional MODIS LCLU product for this African region alone were lower than that of the global MODIS LCLU product overall accuracy (63–66% vs. 75%). However, for countries with uniform or homogenous land cover, the overall accuracy was much higher than the global accuracy and as high as 87% and 78% for Botswana and Namibia, respectively. In addition, the wetland and grassland classes had the highest user’s accuracies in most of the countries (89%–99%), which are the ones with the highest number of MODIS land cover classification algorithm training sites.Our LCLU change analysis revealed that Botswana’s most significant changes were the net reforestation, net grass loss and net wetland expansion. For Rwanda, although there have been significant forest, grass and crop expansions in some areas, there also have been significant forest, grass and crop loss in other areas that resulted in very minimal net changes. As for Tanzania, its most significant changes were the net deforestation and net crop expansion. Malawi’s most significant changes were the net deforestation, net crop expansion, net grass expansion and net wetland loss. Finally, Namibia’s most significant changes were the net deforestation and net grass expansion.The only noticeable environmental driver was in Malawi, which had a significant net wetland loss and could be due to the fact that it was the only country that had a reduction in total precipitation between the periods when the LCLU maps were developed. Not only that, but Malawi also happened to have a slight increase in temperature, which would cause more evaporation and net decrease in wetlands if the precipitation didn’t increase as was the case in that country. In addition, within our studied countries, forestland expansion and loss as well as crop expansion and loss were happening in the same country almost equally in some cases. All of that implies that non-environmental factors, such as socioeconomics and governmental policies, could have been the main drivers of these LCLU changes in many of these countries in E&S Africa. It will be important to further study in the future the detailed effects of such drivers on these LCLU changes in this part of the world.     

Quantifying annual land-cover change and vegetation greenness variation in a coastal ecosystem using dense time-series Landsat data

Land-cover change may affect water and carbon cycles when transitioning from one land-cover category to another (land-cover conversion, LCC) or when the characteristics of the land-cover type are altered without changing its overall category (land-cover modification, LCM). Given the increasing availability of time-series remotely sensed data for earth monitoring, there has been increased recognition of the importance of accounting for both LCC and LCM to study annual land-cover changes. In this study, we integrated 1,513 time-series Landsat images and a change-updating method to identify annual LCC and LCM during 1986–2015 in the coastal area of Zhejiang Province, China. The purpose was to quantify their contributions to land-cover changes and impacts on the amount of vegetation. The results show that LCC and LCM can be successfully distinguished with an overall accuracy of 90.0%. LCM accounted for 22% and 40.5% of the detected land-cover changes in reclaimed and inland areas, respectively, during 1986–2015. In the reclaimed area, LCC occurred mostly in muddy tidal flats, construction land, aquaculture ponds, and freshwater herbaceous land, whereas LCM occurred mostly in freshwater herbaceous land, Spartina alterniflora, and muddy tidal flats. In the inland area, both LCC and LCM were concentrated in forest and dryland. Overall, LCC had a mean magnitude of normalized difference vegetation index (NDVI) change similar to that of LCM. However, LCC had a positive effect and LCM had a negative effect on NDVI change in the reclaimed area. Both LCC and LCM in the inland area had negative impacts on vegetation greenness, but LCC resulted in larger NDVI change magnitude. Impacts of LCC and LCM on vegetation greenness were quantified for each land-cover type. This study provided a methodological framework to take both LCC and LCM into account when analyzing land-cover changes and quantified their effects on coastal ecosystem vegetation.     

利用天宫一号和Landsat7对地观测数据的森林变化检测

由于自然演替和一些干扰因素的影响,森林覆盖处在不断的变化中.结合云南省西双版纳地区的天宫一号高光谱数据以及Landsat影像,研究了热带森林覆盖制图与变化检测的自动化识别方法.首先分析了每景影像中红光波段的光谱属性,依据直方图提取出纯净森林像元,然后计算影像中各像元与纯净森林像元之间的光谱相似性,从而得到森林指数并以此为依据提取出每景影像对应的森林覆盖图,将多期的森林覆盖专题图进行叠加分析即可得到森林变化专题图.结果表明:(1)使用天宫一号高光谱影像可以进行森林覆盖自动化提取,生成的森林覆盖图合理地反映了森林分布状况;(2)与多期遥感影像结合进行森林变化信息提取,提取结果很好地体现了森林减少和森林恢复情况,对新恢复的未郁闭森林也可以进行有效检测.     

Land cover dynamics monitoring with Landsat data in Kunming,China: a cost-effective sampling and modelling scheme using Google Earth imagery and random forests

Changes in forest composition impact ecological services, and are considered important factors driving global climate change. A hybrid sampling method along with a modelling approach to map current and past land cover in Kunming, China is reported. MODIS land cover (2001–2011) data-sets were used to detect pixels with no apparent change. Around 3000 ‘no change points’ were systematically selected and sampled using Google Earth’s high-resolution imagery. Thirty-five per cent of these points were verified and used for training and validation. We used Random forests to classify multi-temporal Landsat imagery. Results show that forest cover has had a net decrease of 14385?ha (1.3% of forest area), which was primary converted to shrublands (11%), urban and barren land (2.7%) and agriculture (2.5%). Our validation indicates an overall accuracy (Kappa) of 82%. Our methodology can be used to consistently map the dynamics of land cover change in similar areas with minimum costs.     

Land cover mapping in the Upper Shire River catchment in Malawi using Landsat satellite data

Land cover mapping forms a reference base for resource managers in their decision-making processes to guide rural/urban growth and management of natural resources. The aim of this study was to map land cover dynamics within the Upper Shire River catchment, Malawi. The article promotes innovation of automated land cover mapping based on remote sensing information to generate data products that are both appropriate to, and usable within different scientific applications in developing countries such as Malawi. To determine land cover dynamics, 1989 and 2002 Landsat images were used. Image bands were combined in transformations and indices with physical meaning; together with spatial data, to enhance classification accuracy. A maximum likelihood classification for each image was computed for identification of land cover variables. The results showed that the combination of spatial and digital data enhanced classification accuracy and the ability to categorise land cover features, which are relatively inhomogeneous.     

Selection and quality assessment of Landsat data for the North American forest dynamics forest history maps of the US

Using the NASA Earth Exchange platform, the North American Forest Dynamics (NAFD) project mapped forest history wall-to-wall, annually for the contiguous US (1986–2010) using the Vegetation Change Tracker algorithm. As with any effort to identify real changes in remotely sensed time-series, data gaps, shifts in seasonality, misregistration, inconsistent radiometry and cloud contamination can be sources of error. We discuss the NAFD image selection and processing stream (NISPS) that was designed to minimize these sources of error. The NISPS image quality assessments highlighted issues with the Landsat archive and metadata including inadequate georegistration, unreliability of the pre-2009 L5 cloud cover assessments algorithm, missing growing-season imagery and paucity of clear views. Assessment maps of Landsat 5–7 image quantities and qualities are presented that offer novel perspectives on the growing-season archive considered for this study. Over 150,000+ Landsat images were considered for the NAFD project. Optimally, one high quality cloud-free image in each year or a total of 12,152 images would be used. However, to accommodate data gaps and cloud/shadow contamination 23,338 images were needed. In 220 specific path-row image years no acceptable images were found resulting in data gaps in the annual national map products.     

Updating Landsat time series of surface-reflectance composites and forest change products with new observations

The use of time series satellite data allows for the temporally dense, systematic, transparent, and synoptic capture of land dynamics over time. Subsequent to the opening of the Landsat archive, several time series approaches for characterizing landscape change have been developed, often representing a particular analytical time window. The information richness and widespread utility of these time series data have created a need to maintain the currency of time series information via the addition of new data, as it becomes available. When an existing time series is temporally extended, it is critical that previously generated change information remains consistent, thereby not altering reported change statistics or science outcomes based on that change information. In this research, we investigate the impacts and implications of adding additional years to an existing 29-year annual Landsat time series for forest change. To do so, we undertook a spatially explicit comparison of the 29 overlapping years of a time series representing 1984–2012, with a time series representing 1984–2016. Surface reflectance values, and presence, year, and type of change were compared. We found that the addition of years to extend the time series had minimal effect on the annual surface reflectance composites, with slight band-specific differences (r ≥ 0.1) in the final years of the original time series being updated. The area of stand replacing disturbances and determination of change year are virtually unchanged for the overlapping period between the two time-series products. Over the overlapping temporal period (1984–2012), the total area of change differs by 0.53%, equating to an annual difference in change area of 0.019%. Overall, the spatial and temporal agreement of the changes detected by both time series was 96%. Further, our findings suggest that the entire pre-existing historic time series does not need to be re-processed during the update process. Critically, given the time series change detection and update approach followed here, science outcomes or reports representing one temporal epoch can be considered stable and will not be altered when a time series is updated with newly available data.     

Generation and evaluation of gross primary productivity using Landsat data through blending with MODIS data

Spatial and Temporal Adaptive Reflectance Fusion Model (STARFM) has been used for the blending of Landsat and MODIS data. Specifically, the 30 m Landsat-7 ETM+ (Enhanced Thematic Mapper plus) surface reflectance was predicted for a period of 10 years (2000–2009) as the product of observed ETM+ and MODIS surface reflectance (MOD09A1) on the predicted and observed ETM+ dates. A pixel based analysis for six observed ETM+ dates covering winter and summer crops showed that the prediction method was more accurate for NIR band (mean r² = 0.71, p ≤ 0.01) compared to green band (mean r² = 0.53; p ≤ 0.01). A recently proposed chlorophyll index (CI), which involves NIR and green spectral bands, was used to retrieve gross primary productivity (GPP) as the product of CI and photosynthetic active radiation (PAR). The regression analysis of GPP derived from closet observed and synthetic ETM+ showed a good agreement (r² = 0.85, p ≤ 0.01 and r² = 0.86, p ≤ 0.01) for wheat and sugarcane crops, respectively. The difference between the GPP derived from synthetic and observed ETM+ (prediction residual) was compared with the difference in GPP values from observed ETM+ on the two dates (temporal residual). The prediction residuals (mean value of 1.97 g C/m² in 8 days) was found to be significantly lower than the temporal residuals (mean value of 4.46 g C/m² in 8 days) that correspondence to 12% and 27%, respectively, of GPP values (mean value of 16.53 g C/m² in 8 days) from observed ETM+ data, implying that the prediction method was better than temporal pixel substitution. Investigating the trend in synthetic ETM+ GPP values over a growing season revealed that phenological patterns were well captured for wheat and sugarcane crops. A direct comparison between the GPP values derived from MODIS and synthetic ETM+ data showed a good consistency of the temporal dynamics but a systematic error that can be read as bias (MODIS GPP over estimation). Further, the regression analysis between observed evapotranspiration and synthetic ETM+ GPP showed good agreement (r² = 0.66, p ≤ 0.01).     

Automatic and improved radiometric correction of Landsat imagery using reference values from MODIS surface reflectance images

Radiometric correction is a prerequisite for generating high-quality scientific data, making it possible to discriminate between product artefacts and real changes in Earth processes as well as accurately produce land cover maps and detect changes. This work contributes to the automatic generation of surface reflectance products for Landsat satellite series. Surface reflectances are generated by a new approach developed from a previous simplified radiometric (atmospheric + topographic) correction model. The proposed model keeps the core of the old model (incidence angles and cast-shadows through a digital elevation model [DEM], Earth–Sun distance, etc.) and adds new characteristics to enhance and automatize ground reflectance retrieval. The new model includes the following new features: (1) A fitting model based on reference values from pseudoinvariant areas that have been automatically extracted from existing reflectance products (Terra MODIS MOD09GA) that were selected also automatically by applying quality criteria that include a geostatistical pattern model. This guarantees the consistency of the internal and external series, making it unnecessary to provide extra atmospheric data for the acquisition date and time, dark objects or dense vegetation. (2) A spatial model for atmospheric optical depth that uses detailed DEM and MODTRAN simulations. (3) It is designed so that large time-series of images can be processed automatically to produce consistent Landsat surface reflectance time-series. (4) The approach can handle most images, acquired now or in the past, regardless of the processing system, with the exception of those with extremely high cloud coverage. The new methodology has been successfully applied to a series of near 300 images of the same area including MSS, TM and ETM+ imagery as well as to different formats and processing systems (LPGS and NLAPS from the USGS; CEOS from ESA) for different degrees of cloud coverage (up to 60%) and SLC-off. Reflectance products have been validated with some example applications: time series robustness (for a pixel in a pseudoinvariant area, deviations are only 1.04% on average along the series), spectral signatures generation (visually coherent with the MODIS ones, but more similar between dates), and classification (up to 4 percent points better than those obtained with the original manual method or the CDR products). In conclusion, this new approach, that could also be applied to other sensors with similar band configurations, offers a fully automatic and reasonably good procedure for the new era of long time-series of spatially detailed global remote sensing data.     

Geospatial method for computing supplemental multi-decadal US coastal land use and land cover classification products,using Landsat data and C-CAP products

This paper discusses the development and implementation of a method that can be used with multi-decadal Landsat data for computing general coastal US land use and land cover (LULC) maps consisting of seven classes. With Mobile Bay, Alabama as the study region, the method that was applied to derive LULC products for nine dates across a 34-year time span. Classifications were computed and refined using decision rules in conjunction with unsupervised classification of Landsat data and Coastal Change and Analysis Program value-added products. Each classification’s overall accuracy was assessed by comparing stratified random locations to available high spatial resolution satellite and aerial imagery, field survey data and raw Landsat RGBs. Overall classification accuracies ranged from 83 to 91% with overall κ statistics ranging from 0.78 to 0.89. Accurate classifications were computed for all nine dates, yielding effective results regardless of season and Landsat sensor. This classification method provided useful map inputs for computing LULC change products.     

Time-series analysis of multi-resolution optical imagery for quantifying forest cover loss in Sumatra and Kalimantan,Indonesia

Monitoring loss of humid tropical forests via remotely sensed imagery is critical for a number of environmental monitoring objectives, including carbon accounting, biodiversity, and climate modeling science applications. Landsat imagery, provided free of charge by the U.S. Geological Survey Center for Earth Resources Observation and Science (USGS/EROS), enables consistent and timely forest cover loss updates from regional to biome scales. The Indonesian islands of Sumatra and Kalimantan are a center of significant forest cover change within the humid tropics with implications for carbon dynamics, biodiversity maintenance and local livelihoods. Sumatra and Kalimantan feature poor observational coverage compared to other centers of humid tropical forest change, such as Mato Grosso, Brazil, due to the lack of ongoing acquisitions from nearby ground stations and the persistence of cloud cover obscuring the land surface. At the same time, forest change in Indonesia is transient and does not always result in deforestation, as cleared forests are rapidly replaced by timber plantations and oil palm estates. Epochal composites, where single best observations are selected over a given time interval and used to quantify change, are one option for monitoring forest change in cloudy regions. However, the frequency of forest cover change in Indonesia confounds the ability of image composite pairs to quantify all change. Transient change occurring between composite periods is often missed and the length of time required for creating a cloud-free composite often obscures change occurring within the composite period itself. In this paper, we analyzed all Landsat 7 imagery with <50% cloud cover and data and products from the Moderate Resolution Imaging Spectroradiometer (MODIS) to quantify forest cover loss for Sumatra and Kalimantan from 2000 to 2005. We demonstrated that time-series approaches examining all good land observations are more accurate in mapping forest cover change in Indonesia than change maps based on image composites. Unlike other time-series analyses employing observations with a consistent periodicity, our study area was characterized by highly unequal observation counts and frequencies due to persistent cloud cover, scan line corrector off (SLC-off) gaps, and the absence of a complete archive. Our method accounts for this variation by generating a generic variable space. We evaluated our results against an independent probability sample-based estimate of gross forest cover loss and expert mapped gross forest cover loss at 64 sample sites. The mapped gross forest cover loss for Sumatra and Kalimantan was 2.86% of the land area, or 2.86 Mha from 2000 to 2005, with the highest concentration having occurred in Riau and Kalimantan Tengah provinces.     

Estimating fractional green vegetation cover of Mongolian grasslands using digital camera images and MODIS satellite vegetation indices

ABSTRACT

Fractional green vegetation cover (FVC) is a useful indicator for monitoring grassland status. Satellite imagery with coarse spatial but high temporal resolutions has been preferred to monitor seasonal and inter-annual FVC dynamics in wide geographic area such as Mongolian steppe. However, the coarse spatial resolution can cause a certain uncertainty in the satellite-based FVC estimation, which calls attention to develop a robust statistical test for the relationship between field FVC and satellite-derived vegetation indices. In the arid and semi-arid Mongolian steppe, nadir pointing digital camera images (DCI) were collected and used to produce a FVC dataset to support the evaluation of satellite-based FVC retrievals. An optimal DCI processing method was determined with respect to three color spaces (RGB, HIS, L*a*b*) and six green pixel classification algorithms, from which a country-wide dataset of DCI-FVC was produced and used for evaluating the accuracy of satellite-based FVC estimates from MODIS vegetation indices. We applied three empirical and three semi-empirical MODIS-FVC retrieval models. DCI data were collected from 96 sites across the Mongolian steppe from 2012 to 2014. The histogram algorithm using the hue (H) value of the HIS color space was the optimal DCI method (r² = 0.94, percent root-mean-square-error (RMSE) = 7.1%). For MODIS-FVC retrievals, semi-empirical Baret model was the best-performing model with the highest r² (0.69) and the lowest RMSE (49.7%), while the lowest MB (+1.1%) was found for the regression model with normalized difference vegetation index (NDVI). The high RMSE (>50% or so) is an issue requiring further enhancement of satellite-based FVC retrievals accounting for key plant and soil parameters relevant to the Mongolian steppe and for scale mismatch between sampling and MODIS data.     

Quantifying shoreline change in Funafuti Atoll,Tuvalu using a time series of Quickbird,Worldview and Landsat data

Funafuti Atoll, Tuvalu is located in the southwestern Pacific Ocean, which has experienced some of the highest rates of global sea-level rise over the past 60 years. Atoll islands are low-lying accumulations of reef-derived sediment that provide the only habitable land in Tuvalu, and are considered vulnerable to the myriad possible impacts of climate change, especially sea-level rise. This study examines the shoreline change of twenty-eight islands in Funafuti Atoll between 2005 and 2015 using 0.65 m QuickBird, 0.46 m WorldView-2, and 0.31 m WorldView-3 imagery using an image segmentation and decision tree classification. Shoreline change estimates are compared to previous study that used a visual interpretation approach. The feasibility of estimating island area with Landsat-8 Operational Land Imager (OLI) data is explored using CLASlite software. Results indicate a 0.13% (0.35 ha) decrease in net island area over the study time period, with 13 islands decreasing in area and 15 islands increasing in area. Substantial decreases in island area occurred on the islands of Fuagea, Tefala and Vasafua, which coincides with the timing of Cyclone Pam in March, 2015. Comparison between the WorldView-2 shoreline maps and those created from Landstat-8 indicate that the estimates tend to be in higher agreement for islands that have an area > 0.5 ha, a compact shape, and no built structures. Ten islands had > 90% agreement, with percent disagreements ranging from 2.78 to 100%. The methods and results of this study speak to the potential of automated EoV shoreline monitoring through segmentation and classification tree approach, which would reduce down data processing and analysis time. With the growing constellation of high and medium spatial resolution satellite-based sensors and the development of semi or fully automated image processing technology, it is now possible to remotely assess the short and medium-term shoreline dynamics on dynamic atolls. Landsat estimates were reasonably matched to those derived from fine resolution imagery, with some caveats about island size and shape.     

Major forest changes and land cover transitions based on plant functional types derived from the ESA CCI Land Cover product

Land use and land cover change are of prime concern due to their impacts on CO₂ emissions, climate change and ecological services. New global land cover products at 300 m resolution from the European Space Agency (ESA) Climate Change Initiative Land Cover (CCI LC) project for epochs centered around 2000, 2005 and 2010 were analyzed to investigate forest area change and land cover transitions. Plant functional types (PFTs) fractions were derived from these land cover products according to a conversion table. The gross global forest loss between 2000 and 2010 is 172,171 km², accounting for 0.6% of the global forest area in year 2000. The forest changes are mainly distributed in tropical areas such as Brazil and Indonesia. Forest gains were only observed between 2005 and 2010 with a global area of 9844 km², mostly from crops in Southeast Asia and South America. The predominant PFT transition is deforestation from forest to crop, accounting for four-fifths of the total increase of cropland area between 2000 and 2010. The transitions from forest to bare soil, shrub, and grass also contributed strongly to the total areal change in PFTs. Different PFT transition matrices and composition patterns were found in different regions. The highest fractions of forest to bare soil transitions were found in the United States and Canada, reflecting forest management practices. Most of the degradation from grassland and shrubland to bare soil occurred in boreal regions. The areal percentage of forest loss and land cover transitions generally decreased from 2000–2005 to 2005–2010. Different data sources and uncertainty in the conversion factors (converting from original LC classes to PFTs) contribute to the discrepancy in the values of change in absolute forest area.     

The effect of atmospheric and topographic correction on pixel-based image composites: Improved forest cover detection in mountain environments

Quantification of forest cover is essential as a tool to stimulate forest management and conservation. Image compositing techniques that sample the most suited pixel from multi-temporal image acquisitions, provide an important tool for forest cover detection as they provide alternatives for missing data due to cloud cover and data discontinuities. At present, however, it is not clear to which extent forest cover detection based on compositing can be improved if the source imagery is firstly corrected for topographic distortions on a pixel-basis. In this study, the results of a pixel compositing algorithm with and without preprocessing topographic correction are compared for a study area covering 9 Landsat footprints in the Romanian Carpathians based on two different classifiers: Maximum Likelihood (ML) and Support Vector Machine (SVM). Results show that classifier selection has a stronger impact on the classification accuracy than topographic correction. Finally, application of the optimal method (SVM classifier with topographic correction) on the Romanian Carpathian Ecoregion between 1985, 1995 and 2010 shows a steady greening due to more afforestation than deforestation.     

Pre-processing of a sample of multi-scene and multi-date Landsat imagery used to monitor forest cover changes over the tropics

In support to the Remote Sensing Survey of the global Forest Resource Assessment 2010, the TREES-3 project has processed more than 12,000 Landsat TM and ETM+ data subsets systematically distributed over the tropics. The project aims at deriving area estimates of tropical forest cover change for the periods 1990-2000-2005. The paper presents the pre-processing steps applied in an operational and robust manner to this large amount of multi-date and multi-scene imagery: conversion to top-of-atmosphere reflectance, cloud and cloud shadow detection, haze correction and image radiometric normalization. The results show that the haze correction algorithm has improved the visual appearance of the image and significantly corrected the digital numbers for Landsat visible bands, especially the red band. The impact of the normalization procedures (forest normalization and relative normalization) was assessed on 210 image pairs: in all cases the correlation between the spectral values of the same land cover in both images was improved. The developed automatic pre-processing chain provided a consistent multi-temporal data set across the tropics that will constitute the basis for an automatic object-based supervised classification.