Article

Reconstructing long term annual deforestation dynamics in Pará and Mato Grosso using the Landsat archive

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

Remote sensing based monitoring of deforestation in the tropics is crucial to better understand global land use change and related changes in ecosystem service provision and to inform governments and civil society on the effectiveness their forest protection policies. In Brazil, deforestation has been closely coupled to the expansion of grazing and cropping systems primarily in the tropical forest, but no spatially explicit high resolution database on deforestation exists that captures tropical forest clearing prior to 2000. The open Landsat archive provides >45 years of imagery and is well suited for wall-to-wall assessments of historic deforestation dynamics which are valuable to policy development and environmental impact assessments. Image analysis procedures for reconstructing long-term deforestation dynamics over large areas need to cope with regions and time periods for which the archive contains heterogeneous data densities on a yearly and decadal basis. We create for the first time yearly 30 m maps of long term, annual deforestation dynamics (LTAD) covering the period from 1984 to 2014 for Pará and Mato Grosso, two Brazilian federal states that cover much of the Brazilian arc-of-deforestation. Our results provide valuable insights into historic deforestation trends, with annually increasing deforestation from 1990 to 1999 for both Pará and Mato Grosso. Peak deforestation occurred in 2004 after which deforestation leveled off – with a more pronounced decrease in Mato Grosso than in Pará. Contrary to Mato Grosso, Pará again experienced increasing annual forest clearing in recent years. For the time period after 2000, we provide an in-depth comparison with two widely used products, the Brazilian PRODES and the Global Forest Change maps (GFC, Hansen et al., 2013). Our deforestation estimates (407,000 ± 42,000 km² at 95% confidence level) are above those provided by PRODES, while GFC results are closer to our estimates for the comparison period. Recent PRODES estimates are consistently below our and the GFC results. Overall, our results exemplify the potential of open image archives for multi-decadal, wall-to-wall and fine grain reconstruction of forest change. The presented approach prototypes similar assessments of tropical forest dynamics globally faced with issues of data scarcity.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... Image composites are gap-free wall-to-wall mosaics using all images from a user-defined study area in a selected time period. Metrics are statistical summaries of spectral values or indices that represent the phenological fingerprint of different land covers during that period (Frantz et al. 2017;Griffiths, Jakimow & Hostert 2018). For our study, we selected all available Landsat imagery for each year from 1985-2016, masked out cloud and cloud shadows, and calculated a set of in total seven metrics (i.e., mean, median, standard deviation, 10th-, 25th-, 75th-and 90th-percentile) for each of Landsat's six multispectral bands, as well as a set of spectral indices (i.e., NBR, NDMI, EVI, MSAVI and the three tasseled cap components "brightness", "greenness", "wetness"). ...
... For each of the circular landscapes, we mapped woodland cover for each year between 1985 and 2016 based on Landsat composite metrics derived at a spatial resolution of 30-m in Google Earth Engine(Gorelick et al. 2017). We used an extensive database of training samples) and hand-://guyra.org.py/informedeforestacion). We used these training data to parameterize a time-calibrated random forest classifier, and classified 31 annual woodland loss maps between 1985 and 2016(Griffiths, Jakimow & Hostert 2018). Each map used satellite data from that year and the previous year, to ensure consistency between years. ...
... Griffiths, Jakimow & Hostert 2018) containing three classes: (1) stable woodland, (2) stable non woodland, (3) woodland loss, and extracted the Landsat metrics for the corresponding years. For example, in case of points sampled into deforestation polygons for the year 2015, we extracted Landsat metrics for the year 2014 (i.e., pre-deforestation) and 2015 (i.e., postdeforestation), and used the same years for our stable classes. ...
Thesis
Full-text available
Landnutzungswandel ist eine der Hauptursachen von Biodiversitätsverlust. In den Tropen und Subtropen führt eine Ausweitung von Agrarflächen zu vermehrter Abholzung der Wälder. Selbst wenn zukünftige Waldrodungen vermieden werden können, ist ein weiterer Artenrückgang sehr wahrscheinlich, da viele Arten zeitverzögert auf Veränderungen reagieren. Die Hauptziele dieser Arbeit waren die Auswirkungen vergangener und aktueller Landnutzung auf Biodiversität im argentinischen Chaco besser zu verstehen und Ansätze zu entwickeln, um negative Effekte schon vor einem lokalen Aussterben zu erkennen. Der argentinische Chaco ist aufgrund seiner Landnutzungsgeschichte, den hohen Abholzungsraten und der hohen Biodiversität bestens für eine solche Untersuchung geeignet. Meine Arbeit zeigt, dass der Artenreichtum an Vögeln und Säugetieren stark durch vergangene Landschaftsmuster beeinflusst wurde, was auf zeitverzögerte Reaktionen auf Landnutzungswandel hindeutet, sowie darauf, dass ein Teil der momentan vorkommenden Arten durch vergangene Landnutzungsänderungen noch aussterben wird. Die zeitverzögerten Reaktionen sind hauptsächlich eine Folge von Lebensraumfragmentierung, mehr noch als von Lebensraumverlust. Meine Ergebnisse zeigen, dass das Vorkommen von Ameisenbären seit 1985 stark rückläufig ist, insbesondere seit 2000, als die Ausweitung von Agrarflächen besonders stark zunahm. Abschließend konnte ich zeigen, dass Pekaris meist in abgelegenen Regionen mit hohem Waldanteil vorkommen, sowie dass physiologischer Stress bei Pekaris negativ mit Nahrungsverfügbarkeit korreliert, jedoch nicht mit Abholzung. Meine Arbeit legt nahe, dass Abholzung generell zum Artensterben im argentinischen Chaco beiträgt. Während manche Arten sehr schnell verschwinden, sterben andere nicht direkt aus, was ein Zeitfenster für Naturschutzmaßnahmen eröffnet. Die hier vorgestellten Ergebnisse können dabei helfen solche Zeitfenster in von Abholzung bedrohten Gebieten zu identifizieren.
... Temporally dense, per-pixel analysis has been parlayed into a wide variety of science and application products. Foundational image datasets built on Landsat best-pixel composites (Griffiths et al., 2013a;Roy et al., 2010;White et al., 2014) have allowed subsequent development of applied maps of condition and change (Gómez et al., 2012;Hansen et al., 2014;Hermosilla et al., 2015aHermosilla et al., , 2015bSchroeder et al., 2011;White et al., 2011;White et al., 2014;Roy, 2014, 2016;Griffiths et al., 2018;Roy and Yan, 2018). Algorithms that focus directly on disturbance have also benefited from the temporally-dense Landsat data (Brooks et al., 2012;Huang et al., 2010;Hughes et al., 2017;Kennedy et al., 2010;, allowing investigation of drivers of change (Alonzo et al., 2016;Pahlevan et al., 2018), national to global-scale assessment of disturbance (Hansen et al., 2013;Fig. ...
... Different disturbance types will have different metric values (e.g., different change magnitudes (m 1 versus m 2 ) and persistances (p 1 versus p 2 )). Hawbaker et al., 2017;Masek et al., 2013;Griffiths et al., 2018) that themselves can play into improved modeled estimates of change (Williams et al., 2016). More broadly, dense time series of Landsat data are eminently suitable for detection of land cover change, including for urban areas (Schneider, 2012), agricultural regions (Hurni et al., 2017;Roy and Yan, 2018), and water dynamics at the global scale (Pekel et al., 2016). ...
... From an algorithm perspective, most compositing approaches applied to Landsat data follow a best-pixel selection strategy, often using simple selection criteria such the maximum NDVI or median NIR (Potapov et al., 2011) or consider multi-band distributions of cloud-free candidate observations Flood, 2013). Other best-pixel approaches additionally consider similarity criteria (Nelson and Steinwand, 2015) or derive a decision through a weighted evaluation of several image and scene based parameters (Griffiths et al., 2013a;White et al., 2014;Frantz et al., 2017;Griffiths et al., 2018). Contrary to best-pixel selection approaches, some approaches calculate new spectral values such as in mean-value compositing (Vancutsem et al., 2007) or generate synthetic images based on harmonic time series fits (Zhu et al., 2015a). ...
Article
Full-text available
Formal planning and development of what became the first Landsat satellite commenced over 50 years ago in 1967. Now, having collected earth observation data for well over four decades since the 1972 launch of Landsat- 1, the Landsat program is increasingly complex and vibrant. Critical programmatic elements are ensuring the continuity of high quality measurements for scientific and operational investigations, including ground systems, acquisition planning, data archiving and management, and provision of analysis ready data products. Free and open access to archival and new imagery has resulted in a myriad of innovative applications and novel scientific insights. The planning of future compatible satellites in the Landsat series, which maintain continuity while incorporating technological advancements, has resulted in an increased operational use of Landsat data. Governments and international agencies, among others, can now build an expectation of Landsat data into a given operational data stream. International programs and conventions (e.g., deforestation monitoring, climate change mitigation) are empowered by access to systematically collected and calibrated data with expected future continuity further contributing to the existing multi-decadal record. The increased breadth and depth of Landsat science and applications have accelerated following the launch of Landsat-8, with significant improvements in data quality. Herein, we describe the programmatic developments and institutional context for the Landsat program and the unique ability of Landsat to meet the needs of national and international programs. We then present the key trends in Landsat science that underpin many of the recent scientific and application developments and followup with more detailed thematically organized summaries. The historical context offered by archival imagery combined with new imagery allows for the development of time series algorithms that can produce information on trends and dynamics. Landsat-8 has figured prominently in these recent developments, as has the improved understanding and calibration of historical data. Following the communication of the state of Landsat science, an outlook for future launches and envisioned programmatic developments are presented. Increased linkages between satellite programs are also made possible through an expectation of future mission continuity, such as developing a virtual constellation with Sentinel-2. Successful science and applications developments create a positive feedback loop—justifying and encouraging current and future programmatic support for Landsat.
... An expansive temporal coverage allows to move from static descriptions of land cover or land use towards the characterization of processes such as long-term changes in surface water ( Pekel et al., 2016), changes in tidal zones ( Murray et al., 2018), or deforestation ( Hansen et al., 2013). In terms of land use intensity, long-term characterization of simple indicators of land cover or land use reveals indicators of cropping intensity ( Estel et al., 2016), cropland abandonment ( Dara et al., 2018) or post-deforestation land use intensity ( Griffiths et al., 2018;Müller et al., 2016;. Landsat provides unique opportunities in this regard, as the Landsat image archive provides more than three decades of global coverage at 30 m spatial resolution (Figure 1.6). ...
... The applicability of individual methods is dependent on the operational sensor constellation during the period of interest. For past decades, highly irregular interannual acquisition densities in the Landsat archive pose challenges for long-term mapping approaches ( Griffiths et al., 2018). Seasonally restricted spectral-temporal metrics are promising tools for mapping land use across large areas in past decades ( Schmidt et al., 2016). ...
... Similar to previous studies, we assumed that Landsat-based spectral-temporal metrics provide temporally consistent means for cropland mapping (Deines et al., 2017;Schmidt et al., 2016). Based on this assumption, we trained a generalized non-parametric classifier to predict spectral-temporal metrics across the study period, similar to a recent study targeting long-term mapping of deforestation ( Griffiths et al., 2018). We could thereby restrict training data collection to a year, for which very-high resolution imagery was available in Google Earth. ...
Thesis
Full-text available
A growing world population, and increasing demands for food, feed, fuel and fiber, substantially add pressure on the global land system. The construction of dams is a common strategy for boosting production outputs through irrigation. Reservoirs represent the most important source of irrigation water globally, but their effects on agricultural land systems are only poorly understood. Remote sensing emerges as a key tool for enabling spatially explicit assessments of dam-induced land system change due to its ability to provide spatially detailed, frequent, and synoptic observations of the land surface. The overall goal of this thesis was to assess the effects of irrigation dams on agricultural land systems on a global and regional scale, by making use of state-of-the art remote sensing data products and methods. A synthesis of the current scientific literature offered primary insights into dam-induced changes in agricultural systems, and raised the hypothesis that irrigation dams caused overall increases in agricultural land use intensity. On a global scale, satellite-based measurements of cropping frequency derived from MODIS-based map products attested to this finding, albeit a strong regional variability was apparent. Landsat-based time series methods were used on a national to regional scale, which further revealed strong spatio-temporal dynamics of irrigated agriculture. The results of this thesis add knowledge and spatially explicit insights on the effects of dams on agricultural land systems. The work further emphasizes the important role of remote sensing technologies in exploring future pathways of agricultural intensification.
... Temporally dense, per-pixel analysis has been parlayed into a wide variety of science and application products. Foundational image datasets built on Landsat best-pixel composites (Griffiths et al., 2013a;Roy et al., 2010;White et al., 2014) have allowed subsequent development of applied maps of condition and change (Gómez et al., 2012;Hansen et al., 2014;Hermosilla et al., 2015aHermosilla et al., , 2015bSchroeder et al., 2011;White et al., 2011;White et al., 2014;Roy, 2014, 2016;Griffiths et al., 2018;Roy and Yan, 2018). Algorithms that focus directly on disturbance have also benefited from the temporally-dense Landsat data (Brooks et al., 2012;Huang et al., 2010;Hughes et al., 2017;Kennedy et al., 2010;, allowing investigation of drivers of change (Alonzo et al., 2016;Pahlevan et al., 2018), national to global-scale assessment of disturbance (Hansen et al., 2013;Fig. ...
... Different disturbance types will have different metric values (e.g., different change magnitudes (m 1 versus m 2 ) and persistances (p 1 versus p 2 )). Hawbaker et al., 2017;Masek et al., 2013;Griffiths et al., 2018) that themselves can play into improved modeled estimates of change (Williams et al., 2016). More broadly, dense time series of Landsat data are eminently suitable for detection of land cover change, including for urban areas (Schneider, 2012), agricultural regions (Hurni et al., 2017;Roy and Yan, 2018), and water dynamics at the global scale (Pekel et al., 2016). ...
... From an algorithm perspective, most compositing approaches applied to Landsat data follow a best-pixel selection strategy, often using simple selection criteria such the maximum NDVI or median NIR (Potapov et al., 2011) or consider multi-band distributions of cloud-free candidate observations Flood, 2013). Other best-pixel approaches additionally consider similarity criteria (Nelson and Steinwand, 2015) or derive a decision through a weighted evaluation of several image and scene based parameters (Griffiths et al., 2013a;White et al., 2014;Frantz et al., 2017;Griffiths et al., 2018). Contrary to best-pixel selection approaches, some approaches calculate new spectral values such as in mean-value compositing (Vancutsem et al., 2007) or generate synthetic images based on harmonic time series fits (Zhu et al., 2015a). ...
Article
Full-text available
Formal planning and development of what became the first Landsat satellite commenced over 50 years ago in 1967. Now, having collected earth observation data for well over four decades since the 1972 launch of Landsat-1, the Landsat program is increasingly complex and vibrant. Critical programmatic elements are ensuring the continuity of high quality measurements for scientific and operational investigations, including ground systems, acquisition planning, data archiving and management, and provision of analysis ready data products. Free and open access to archival and new imagery has resulted in a myriad of innovative applications and novel scientific insights. The planning of future compatible satellites in the Landsat series, which maintain continuity while incorporating technological advancements, has resulted in an increased operational use of Landsat data. Governments and international agencies, among others, can now build an expectation of Landsat data into a given operational data stream. International programs and conventions (e.g., deforestation monitoring, climate change mitigation) are empowered by access to systematically collected and calibrated data with expected future continuity further contributing to the existing multi-decadal record. The increased breadth and depth of Landsat science and applications have accelerated following the launch of Landsat-8, with significant improvements in data quality. Herein, we describe the programmatic developments and institutional context for the Landsat program and the unique ability of Landsat to meet the needs of national and international programs. We then present the key trends in Landsat science that underpin many of the recent scientific and application developments and follow-up with more detailed thematically organized summaries. The historical context offered by archival imagery combined with new imagery allows for the development of time series algorithms that can produce information on trends and dynamics. Landsat-8 has figured prominently in these recent developments, as has the improved understanding and calibration of historical data. Following the communication of the state of Landsat science, an outlook for future launches and envisioned programmatic developments are presented. Increased linkages between satellite programs are also made possible through an expectation of future mission continuity, such as developing a virtual constellation with Sentinel-2. Successful science and applications developments create a positive feedback loop-justifying and encouraging current and future programmatic support for Landsat.
... Soybeans and livestock are the main substitutes for forests, especially with expanding export markets (Arima et al., 2011;Walker et al., 2013;Barbosa et al., 2023). Extreme degradation, with deforestation due to logging, successive droughts, and forest fires, replaces the forest with non-forest open vegetation (Griffiths et al., 2018). This situation has been increasing and will likely continue to grow significantly in the future. ...
... The peak of this process occurred in 2004 during the period analyzed, with more than 220,000 acres open to using permanent and temporary crops. The results attest to the research carried out byGriffiths et al. (2018), where the peak of deforestation in 2004 coincides and, consequently, a sharp drop in the following years. However, this year the conversion was more significant than in 2003 due to the growth of crops in the forest.Overall, the conversion of pasture to crop and forest to crop remained on average in the years analyzed. ...
Article
Full-text available
Deforestation in the Amazon Forest has increased exponentially in recent years. This is a consequence of local, regional, and global dynamic economical processes changing land-use and land cover (LULC) in the Amazon Forest. There is evidence that land deforested to be used for pasture is now being used for crop production. In this circumstance, this paper analyzes the Amazon deforestation by agribusiness production displacement on the southeastern region of the Brazilian Amazon Forest. The investigated area is located in the north of the Brazilian State Mato Grosso. The data was collected from the MapBiomas project in a raster format, from 1985 to 2020, identifying 13 LULC classes, with forests, agriculture, and pasture being the most significant. The study employed map algebra to process spatial and temporal LULC changes, photointerpretation for visual validation, and correlation statistics to explore relationships between deforestation, pasture, agriculture, and fires. The results found a strong correlation (0.98) between deforestation and conversion of forest to pasture, and moderate correlation (0.59) between forest to crop conversion and deforestation. Over the last 20 years, 59,663.51 km² of native forests were lost, primarily converted into pasture (17,047.27 km²) and agriculture (42,034.18 km²). Indigenous territories showed minimal deforestation compared to non-demarcated areas. A historical analysis of policies on deforestation in Brazil, and in the State of Mato Grosso, was carried out, demonstrating that in recent years there has been decreased control over the issue. This study provides insights for policymakers to leverage the global policy window for Amazon conservation under the Paris Agreement.
... The matrix shows that the forest clearing class had a moderate rate of commission errors (user's accuracy of 76.9%), which means areas of no change were counted as forest cover change on a 23.1% rate, while it shows a lower rate of errors of omission (producer's accuracy of 83%). We considered the results satisfactory for showing overall forest clearing trends spatially and temporally and comparable to other studies at a similar geographic scale [28,29]. The estimated error-adjusted area of detected annual forest clearings from the year 2000 until the year 2021 was 230,511 ha in total (±19,979 ha). ...
... overall forest clearing trends spatially and temporally and comparable to other studies at a similar geographic scale[28,29]. ...
Article
Full-text available
The Puuc Biocultural State Reserve (PBSR) is a unique model for tropical dry forest conservation in Mexico. Preserving forest biodiversity and carbon within the PBSR depends on maintaining low-impact productive activities coordinated by multiple communal and private landowners. In this study, we used state-of-the-art remote sensing data to investigate past spatial patterns in forest clearing dynamics and their relation to forest carbon density and forest plant species richness and diversity in the context of the forest conservation goals of the PBSR. We used a Landsat-based continuous change detection product for the 2000–2021 period and compared it to carbon density and tree species richness models generated from ALOS-2 PALSAR 2 imagery and national scale forest inventory data. The estimated error-adjusted area of detected annual forest clearings from the year 2000 until the year 2021 was 230,511 ha in total (�19,979 ha). The analysis of annual forest clearing frequency and area suggests that although forest clearing was significantly more intensive outside of the PBSR than within the PBSR during the entire 2000–2021 period, there is no evidence suggesting that the frequency and magnitude of forest clearing changed over the years after the creation of the PBSR in 2011. However, an emergent hotspot analysis shows that high spatiotemporal clustering of forest clearing events (hotspots) during the 2012–2021 period was less common than prior to 2011, and these more recent hotspots have been confined to areas outside the PBSR. After comparing forest clearing events to carbon density and tree species richness models, the results show that landowners outside the PBSR often clear forests with lower carbon density and species diversity than those inside the PBSR. This suggests that, compared to landowners outside the PBSR, landowners within the PBSR might be practicing longer fallow periods allowing forests to attain higher carbon density and tree species richness and hence better soil nutrient recovery after land abandonment. In conclusion, our results show that the PBSR effectively acted as a stabilizing forest management scheme during the 2012–2021 period, minimizing the impact of productive activities by lowering the frequency of forest clearing events and preserving late secondary forests within the PBSR. We recommend continuing efforts to provide alternative optimal field data collection strategies and modeling techniques to spatially predict key tropical forest attributes. Combining these models with continuous change detection datasets will allow for underlying ecological processes to be revealed and the generation of information better adapted to forest governance scales.
... We then applied a minimum mapping unit of 1 ha for the two agriculture classes and others. Deforested areas were only allowed to occur in primary forests, i.e. forests never mapped as deforested in Griffiths et al. (2018) or a previous BLCM. Removal of secondary natural vegetation was labeled as agriculture or others. ...
... We applied a minimum mapping unit of 3 ha to reduce commission errors. Finally, we assessed for each deforestation pixel the year of deforestation according based on our BLCMs and the land-cover maps from Griffiths et al. (2018). ...
Article
Full-text available
The increasing deforestation and fires since 2019 raises concerns about the irreversible destruction of the Brazilian Amazon. Our goal was to better understand these changes in south-west Pará across different land-tenure and farm systems and between the terms of President Rousseff, Temer, and Bolsonaro. We reconstructed deforestation and fire history using all Landsat and Sentinel-2 observations from 2014 to 2020 and assessed, using quasi-experimental methods, the average treatment effects of each presidency on deforestation and fires across land-tenure and farm types. Deforestation nearly quadrupled to 1,201 km², particularly during Bolsonaro in undesignated areas and conservation units and on medium-sized farms (p < 0.001). Burning increased to 4,805 km² and in all tenure types (p < 0.001). The increase was strongest in agrarian settlements and conservation units and on medium and large farms. Our observations show the importance of clarifying land-tenure and re-strengthening disincentives of environmental infractions, which have been weakened specifically under President Bolsonaro.
... These results agree with the general observation that annual rates of changes are typically low when estimated over large areas. For example, for the Brazilian federal state of Para that includes our study site [28], the mean annual deforestation rate using 30-m Landsat long-term time series for the period spanning from 1984 to 2014 was estimated to be 0.66%. ...
... While most existing land cover change assessment approaches either focus specifically on the change using multi-year classification methods (e.g. [14], [15], [28] or rely on a validation strategy to adjust the derived change area estimate [23], the proposed HMM framework allows us to take advantage of the mapping exercise completed independently at regular time intervals. ...
Article
This article aims at investigating the hidden Markov model (HMM) approach for the automated processing of classified satellite images for land cover and land-use change (LCLUC). HMM’s account for transitions between classes at the same location, but that cannot be directly observed due to classification errors. Using a set of transition and emission probabilities, HMM’s allow filtering out errors and recovering the actual sequence of LCLUC, which are typically overestimated when directly estimated from the classified images. After presenting the HMM framework, the methodology is illustrated on three 300-m annual time series of classified images from 2003 to 2019 over 756×756756\times756 km 2 areas in Brazil, People’s Republic of China, and Mali. It is shown how the emission and transition probabilities can be estimated from these time series using a simple Viterbi training, alleviating computationally demanding algorithms. Special attention is paid to the processing of missing observations caused by clouds. Combining these three datasets with a simulation study, it is concluded that the HMM emission and transition probabilities can be estimated with low biases and variances thanks to the vast number (hundreds of thousands) of pixels at hand. The speed of the Viterbi training and decoding steps makes it possible to consider large-scale land cover mapping at moderate or even high spatial resolution as long as the legend of the LCLUC involves a reasonable number of classes like the six main Intergovernmental Panel on Climate Change (IPCC) land categories.
... Dentro desse contexto, o estado de Mato Grosso, em especial, a região de transição Amazônia/Cerrado, merece uma atenção especial em termos de monitoramento da dinâmica de uso e cobertura de terras, pois ela se constitui em uma frente de expansão agrícola para o interior da Amazônia brasileira, por exemplo, para as regiões de Novo Progresso e Santarém no estado do Pará, localizadas ao longo da rodovia federal BR-163, também conhecida como rodovia Cuiabá-Santarém. Estudos anteriores sobre o monitoramento de uso e cobertura de terras envolvendo corte raso e discriminação de tipos de culturas agrícolas anuais ou de identificação de áreas com degradação florestal por corte seletivo no estado de Mato Grosso têm sido conduzidos principalmente com base na análise de imagens ópticas dos satélites da série Landsat (e.g., SHIMABUKURO et al., 2014;ZHU et al., 2016;JAKIMOW;HOSTERT, 2018) e do sensor Moderate Resolution Imaging Spectroradiometer (MODIS) (e.g., ARVOR et al., 2011;MACEDO et al., 2012;VICTORIA et al., 2012;BROWN et al., 2013;PICOLI et al., 2018). ...
... Dentro desse contexto, o estado de Mato Grosso, em especial, a região de transição Amazônia/Cerrado, merece uma atenção especial em termos de monitoramento da dinâmica de uso e cobertura de terras, pois ela se constitui em uma frente de expansão agrícola para o interior da Amazônia brasileira, por exemplo, para as regiões de Novo Progresso e Santarém no estado do Pará, localizadas ao longo da rodovia federal BR-163, também conhecida como rodovia Cuiabá-Santarém. Estudos anteriores sobre o monitoramento de uso e cobertura de terras envolvendo corte raso e discriminação de tipos de culturas agrícolas anuais ou de identificação de áreas com degradação florestal por corte seletivo no estado de Mato Grosso têm sido conduzidos principalmente com base na análise de imagens ópticas dos satélites da série Landsat (e.g., SHIMABUKURO et al., 2014;ZHU et al., 2016;JAKIMOW;HOSTERT, 2018) e do sensor Moderate Resolution Imaging Spectroradiometer (MODIS) (e.g., ARVOR et al., 2011;MACEDO et al., 2012;VICTORIA et al., 2012;BROWN et al., 2013;PICOLI et al., 2018). ...
Article
Full-text available
A região de transição entre Amazônia e Cerrado, especialmente no estado de Mato Grosso, é bastante sensível em termos ambientais por causa da elevada biodiversidade e alta produção de grãos e carne bovina. O objetivo deste estudo foi discriminar classes representativas de uso e cobertura de terras encontradas na região de Sinop/Mato Grosso com base nas imagens de radar do satélite ALOS-2/PALSAR-2. As seguintes classes temáticas foram consideradas: floresta primária; floresta secundária; cultura agrícola; e pastagem cultivada. As imagens foram obtidas nos meses de fevereiro (estação chuvosa) e setembro (estação seca) de 2016, com resolução espacial de 6,25 metros, polarizações HH e HV e banda L (comprimento de onda de 23 cm), as quais foram processadas pelos algoritmos de classificação supervisionada Random Forest (RF) e Support Vector Machine (SVM). Amostras de treinamento e de validação (65 amostras) foram obtidas em campo e complementadas com base nos mapas de uso e cobertura de terras produzidos pelos projetos MapBiomas e TerraClass Amazônia (135 amostras). Foi possível discriminar dois grupos de classes temáticas: floresta primária e floresta secundária; e cultura agrícola e pastagem cultivada. Apesar da acurácia global do RF ter sido superior ao do SVM, os dois classificadores mostraram desempenhos estatisticamente similares.
... This topic is highly quoted, and discussions have concluded that representative sample datasets make all the difference in the improvement of classifications by different methods [126]. Therefore, the training dataset composition remains an issue, especially in heterogeneous areas with scarce data [127]. Different authors discussed the impact of the low availability of representative samples to perform accurate LULC classification. ...
... In recent years, the exploring of the electromagnetic spectrum by the broader science community has expanded with the advent of Earth Observation missions characterized by global coverage and medium spatial resolution data, especially Landsat-like (10-30 m) missions, as the L8/OLI and the S2/MSI, which jointly have provided global diffusion of narrow and sensitive spectral bands within this range of spatial resolution [28,40]. The most highlighted bands in L8/OLI and S2/MSI are the SWIR and Red-edge, and VIs based on them have increased classification accuracies in different landscapes around the world [85,127,144]. Most of these VIs were little explored for a long time [40] but turned on only after the S2/MSI mission advent. ...
Article
Full-text available
Recent applications of Landsat 8 Operational Land Imager (L8/OLI) and Sentinel-2 MultiSpectral Instrument (S2/MSI) data for acquiring information about land use and land cover (LULC) provide a new perspective in remote sensing data analysis. Jointly, these sources permit researchers to improve operational classification and change detection, guiding better reasoning about landscape and intrinsic processes, as deforestation and agricultural expansion. However, the results of their applications have not yet been synthesized in order to provide coherent guidance on the effect of their applications in different classification processes, as well as to identify promising approaches and issues which affect classification performance. In this systematic review, we present trends, potentialities, challenges, actual gaps, and future possibilities for the use of L8/OLI and S2/MSI for LULC mapping and change detection. In particular, we highlight the possibility of using medium-resolution (Landsat-like, 10-30 m) time series and multispectral optical data provided by the harmonization between these sensors and data cube architectures for analysis-ready data that are permeated by publicizations, open data policies, and open science principles. We also reinforce the potential for exploring more spectral bands combinations, especially by using the three Red-edge and the two Near Infrared and Shortwave Infrared bands of S2/MSI, to calculate vegetation indices more sensitive to phenological variations that were less frequently applied for a long time, but have turned on since the S2/MSI mission. Summarizing peer-reviewed papers can guide the scientific community to the use of L8/OLI and S2/MSI data, which enable detailed knowledge on LULC mapping and change detection in different landscapes, especially in agricultural and natural vegetation scenarios.
... Deriving temporally aggregated, (e.g., seasonal or annual) features from Landsat image time series can help to overcome such issues, and aid in improving the thematic detail, consistency, and quality of maps in agricultural systems [23][24][25]. Common techniques to generate standardized gap-free spectral features include pixel-based compositing [26][27][28], the computation of spectral-temporal metrics [29][30][31], as well as data fusion [32,33] or gap-filling techniques [34]. Such temporal features can be produced consistently for multiple periods and large areas [35]. ...
... The applicability of individual methods is dependent on the operational sensor constellation during the period of interest. For past decades, highly irregular inter-annual acquisition densities in the Landsat archive pose challenges for long-term mapping approaches [28]. Seasonally restricted spectral-temporal metrics are promising tools for mapping land use across large areas in past decades [31]. ...
Article
Full-text available
Spatially explicit information on cropland use intensity is vital for monitoring land and water resource demands in agricultural systems. Cropping practices underlie substantial spatial and temporal variability, which can be captured through the analysis of image time series. Temporal binning helps to overcome limitations concerning operability and repeatability for mapping large areas and can improve the thematic detail and consistency of maps in agricultural systems. We here assessed the use of annual, quarterly, and eight-day temporal features for mapping five cropping practices on annual croplands across Turkey. We used 2,403 atmospherically corrected and topographically normalized Landsat Collection 1 L1TP images of 2015 to compute quarterly best-pixel composites, quarterly and annual spectral-temporal metrics, as well as gap-filled eight-day time series of Tasseled Cap components. We tested 22 feature sets for binary cropland mapping, and subsequent discrimination of five cropping practices: Spring and winter cropping, summer cropping, semi-aquatic cropping, double cropping, and greenhouse cultivation. We evaluated area-adjusted accuracies and compared cropland area estimates at the province-level with official statistics. We achieved overall accuracies above 90%, when using either all quarterly features or the eight-day Tasseled Cap time series, indicating that temporal binning of intra-annual image time-series into multiple temporal features improves representations of cropping practices. Class accuracies of winter and spring, summer, and double cropping were robust, while omission errors for semi-aquatic cropping and greenhouse cultivation were high. Our mapped cropland extent was in good agreement with province-level statistics (r² = 0.85, RMSE = 7.2%). Our results indicate that 71.3% (± 2.3%) of Turkey´s annual croplands were cultivated during winter and spring, 15.8% (± 2.2%) during summer, while 8.5% (± 1.6%) were double-cropped, 4% (± 1.9%) were cultivated under semi-aquatic conditions, and 0.32% (± 0.2%) was greenhouse cultivation. Our study presents an open and readily available framework for detailed cropland mapping over large areas, which bears the potential to inform assessments of land use intensity, as well as land and water resource demands.
... However, their results show limitations when using medium-(30-5) to low-spatial resolution (30-250 m), as those images have limited details to differentiate the vegetation types. Using medium-to-high resolution (5-1 m) imagery makes visual interpretation very expensive and often impractical (Arvor et al., 2021;Griffiths et al., 2018). Machine learning methods such as random forest (RF) (Neto et al., 2016(Neto et al., , 2018Neves et al., 2019;Chaves et al., 2021;Alencar et al., 2020;de Souza Mendes et al., 2019;Lewis et al., 2022;Bendini et al., 2020) and support vector machines (SVM) (Schwieder et al., 2015;Camargo et al., 2019) have been applied to map the vegetation in medium, high and very-high-resolution (VHR) (<1 m) images. ...
... Monitoring deforestation in tropical forests has been addressed by several researchers over the years, using a variety of methods. These approaches include statistical methods, as demonstrated by Schultz et al. (2016), McRoberts & Walters (2012), Smith et al. (2019), and Cabral et al. (2018); spatial modeling methods, exemplified by Griffiths et al. (2018) and Phua et al. (2008); and by machine learning techniques, according to Bragagnolo et al. (2021), Grings et al. (2020), and Khan et al. (2017). These techniques have been successfully applied to detect and monitor changes in forest cover in diverse geographic contexts, such as Malaysia (Phua et al., 2008), the tropics (Schultz et al., 2016), Peru (Tarazona & Miyasiro-López, 2020), Amazônia (da Silva et al., 2019), and the Argentine Chaco region (Grings et al., 2020). ...
Article
Full-text available
In this study, we applied a multivariate logistic regression model to identify deforested areas and evaluate the current effects on environmental variables in the Brazilian state of Rondônia, located in the southwestern Amazon region using data from the MODIS/Terra sensor. The variables albedo, temperature, evapotranspiration, vegetation index, and gross primary productivity were analyzed from 2000 to 2022, with surface type data from the PRODES project as the dependent variable. The accuracy of the models was evaluated by the parameters area under the curve (AUC), pseudo R², and Akaike information criterion, in addition to statistical tests. The results indicated that deforested areas had higher albedo (25%) and higher surface temperatures (3.2 °C) compared to forested areas. There was a significant reduction of the EVI (16%), indicating water stress, and a decrease in GPP (18%) and ETr (23%) due to the loss of plant biomass. The most precise model (91.6%) included only surface temperature and albedo, providing important information about the environmental impacts of deforestation in humid tropical regions.
... Numerous studies have shown that deforestation significantly reduces precipitation and ET (Costa and Foley, 2000;Nóbrega, 2014;Santos et al., 2017;Saddique et al., 2020). Other works carried out in the Amazon Forest highlight the impact of vegetation cover changes on the water balance (Pongratz et al., 2006;Hayhoe et al., 2011;Griffiths et al., 2018;Cabral Júnior et al., 2022) and actual ET (Souza et al., 2019;Kohler et al., 2021;Paiva et al., 2023). ...
Article
Full-text available
Several studies have shown that changes in land cover within a given watershed significantly affect the hydrological cycle and its variables. In the Xingu basin, many areas had their vegetation replaced by agricultural crops and pastures, while deforestation has been particularly prevalent in the region known as the Arch of Deforestation. Using remote sensing techniques enable the estimation of biophysical variable ETr for extensive areas, as exemplified in the study basin. Evapotranspiration data used in this work were obtained by creating a product that returns the combined median of the MOD16A2, PML_V2, Terra Climate, GLEAM_v3.3a, FLUXCOM, SSEBop, FLDAS, and ERA5-Land models, with subsequent application of the data provided by Collection 6 of the MapBiomas network, allowing the integration of land use and land cover information with real evapotranspiration estimates for the transition ranges: Forest to Pasture; Forest to Agricultural Land; Cerrado to Pasture; Cerrado to Agricultural Land. The interval defined for the study corresponds to the years 1985 to 2020, according to the historical series available on MapBiomas. After applying programming languages to filter the data, the results underwent statistical analysis to elucidate the effects of soil changes on evapotranspiration. Over the total data period (1985-2020), there was a decrease in forest areas (-16.23%), with conversion to pasture areas, in the order of +12.51%, and agricultural areas, reaching +5.5%. In the same timeframe, evapotranspiration in conversion bands underwent minimal changes, notably from 2009 to 2020, where a decreasing trend was reported of 0.095 mm/month for the “forest to pasture” substitution, and 0.090 mm/month in “Cerrado for pasture”.
... According to our results, Landsat composite bands and indices derived using percentiles (10 th , 50 th , 90 th , and IQR) were crucial for all the classifications. This evidence confirms previous studies suggesting that band and spectral index statistics can improve the accuracy of land cover classification (Griffiths, Jakimow, and Hostert 2018;Nill et al. 2022;Van De Kerchove et al. 2021) and suggests that these additional indices are frequently more important than the simple, more frequently used median. ...
Article
Full-text available
Understanding grassland habitat dynamics in space and time is crucial for evaluating the effectiveness of protection measures and developing sustainable management practices, specifically within the Natura 2000 network and in light of the European Biodiversity Strategy. Land cover maps, derived from remote sensing data, are essential for understanding long-term changes in vegetation cover and land use and assessing the impact of land use changes on grassland ecosystems. In this study, we conducted a 20-year land cover analysis of grassland landscapes in Umbria, Italy, using Random Forest classifications of Landsat data in Google Earth Engine. Our analysis was based on the years 2000, 2010, and 2020. We integrated harmonic modeling, Gray-Level Co-occurrence Matrix (GLCM) textural analysis, statistical image and gradient analysis, and other spectral and Digital Terrain Model (DTM)-derived indices to enhance the classification capabilities. The LandTrendr (LT) algorithm was used in GEE to collect ground control points in no-change areas automatically. We used a method based on Multilayer Perceptron-Artificial Neural Networks (MLP-ANNs) to forecast 2040 land cover. Our land cover classifications and the scenario model validation achieved an overall accuracy of over 90%. However, the classification of shrublands proved challenging due to their mixed composition and unique spatial patterns, resulting in lower accuracies. Feature importance analysis demonstrated the value of the enhanced map composition, and applying the LandTrendr algorithm simplified the diachronic land use and land cover (LULC) classification and change analysis by supporting automatic training data collection. Results support the interpretation of grassland dynamics in Umbria over the past two decades and identify areas affected by encroachment from shrubs, woody plants, or those with reduced green biomass. The forecasting method along with the selection of spatial drivers to predict land cover change, demonstrated high efficiency compared to other studies. A specific analysis was developed to identify areas where conservation measures related to the Natura 2000 network have been more or less effective in preserving grasslands. Overall, the research provides a scientific foundation for a methodology helpful in informing policy decisions and defining spatially explicit management strategies to enhance grassland conservation inside and outside Natura 2000 areas.
... "We came to Pará because there was plenty of virgin forest left," he says. The situation in Mato Grosso is different: since the mid-1980s, roughly 40% of its rainforest has been cut down 4 . ...
... According to our results, Landsat composite bands and indices derived using percentiles (10 th , 50 th , 90 th , and IQR) were crucial for all the classifications. This evidence confirms previous studies suggesting that band and spectral index statistics can improve the accuracy of land cover classification (Griffiths, Jakimow, and Hostert 2018;Nill et al. 2022;Van De Kerchove et al. 2021) and suggests that these additional indices are frequently more important than the simple, more frequently used median. ...
Chapter
Full-text available
Understanding how habitats in the European Natura 2000 network change over time and space is crucial to evaluating the effectiveness of protective measures and developing sustainable management practices. Satellite remote sensing through Earth observation offers a cost-effective, timely, reproducible vegetation analysis. This study aims to analyze the changes in grassland habitats (types 6210 and 6230*, Annex I, European Directive 92/43/EEC) within the Natura 2000 network’s areas in Umbria (Central Italy) between 2000 and 2020. The goal is to gather spatial information to develop more sustainable planning and management strategies. In this direction, we selected a three-year time window in two periods (1999–2001 and 2019–2021). We applied an enhanced dataset composition, including cloud-filtering, topography correction, and textural analysis on Landsat 7 and 8 surface reflectance images available in Google Earth Engine (GEE). Additional morphometric features were derived from the digital elevation model of the area. Using machine learning classification (Random Forest – RF), we obtained the land cover maps for the two periods under investigation with high overall accuracy (87 and 88%). Only shrublands showed a lower classification accuracy due to the varied nature of this land cover class and the limited resolution of Landsat data. On this basis, we identified transformations that occurred in the study areas. The results confirmed the effectiveness of enhanced map composition and RF in GEE for diachronic land cover classification. The final spatial information can help identify areas where conservation measures related to the Natura 2000 network have been more or less effective and develop more appropriate management strategies for grasslands of European concern.Keywordsgrasslands dynamicsgoogle earth engineremote sensingmachine learningrandom forestLandsat 7Landsat 8
... The matrix shows that the forest clearing class had a moderate rate of commission errors (user's accuracy of 76.9%) which means areas of no change were counted as forest cover change on a 23.1% rate, while it shows a lower rate of errors of omission (producer's accuracy of 83%). We considered the results satisfactory for showing overall forest clearing trends spatially and temporally, and comparable to other studies at a similar geographic scale (Diaz-Gallegos et al. 2010, Griffiths et al. 2018. ...
Preprint
Full-text available
The Puuc Biocultural State Reserve (PBSR is an unique model for tropical dry forest conservation in Mexico. Preserving forest biodiversity and carbon within the PBSR depends on the mainte-nance of low impact productive activities coordinated by multiple communal and private land-owners. In this study, we used state-of-the-art remote sensing data to investigate past spatial patterns in forest clearing dynamics and its relation to forest carbon density and forest plant species richness and diversity in the context of the forest conservation goals of the PBSR. We used a Landsat-based continuous change detection product for the 2000-2021 period and compared it to carbon density and tree species richness models generated from ALOS-2 PALSAR 2 imagery and national scale forest inventory data. The estimated error-adjusted area of detected annual forest clearings from the year 2000 until the year 2021 was 230,511 ha in total (+19,979 ha). The analysis of annual forest clearing frequency and area suggests that although forest clearing was significantly more intensive outside of the PBSR than within the PBSR during the entire 2000-2021 period, there is no evidence suggesting that the frequency and magnitude of forest clearing has changed over the years after the creation of the PBSR in 2011. An emergent hotspot analysis shows, however, that high spatiotemporal clustering of forest clearing events (hotspots) during the 2012-2021 pe-riod was less common than prior to 2011 and these more recent hotspots have been confined to areas outside the PBSR. After comparing forest clearing events to carbon density and tree species richness models, results show that land owners outside the PBSR often clear forests with lower carbon density and species diversity than land owners inside the PBSR. This suggests that, compared to land owners outside the PBSR, land owners within the PBSR might be practicing longer fallow periods allowing forests to attain higher carbon density and tree species richness and hence better soil nutrient recovery after land abandonment. In conclusion, our results show that the PBSR effectively acted as a stabilizing forest management scheme during the 2012-2021 period minimizing the impact of productive activities by lowering the frequency of forest clearing events and preserving late secondary forests within the PBSR. We recommend the continuation of efforts for providing alternative optimal field data collection strategies and modeling techniques to spatially predict key tropical forest attributes. The combination of these models with contin-uous change detection datasets will allow to reveal underlying ecological processes and generate information better adapted to forest governance scales.
... (Diaz-Gallegos et al. 2010, Griffiths et al. 2018.Table 1. Area weighted confusion matrix showing the result of the accuracy assessment performed on the forest baseline map for the year 2000 and the annual forest clearing dataset.The estimated error-adjusted area of detected annual forest clearings from the year 2000 until the year 2021 was 230,511 ha in total (±19,979 ha). Results also show a mean and median forest clearing patch area of 1.26 ha and 0.54 ha respectively (Std Dev = 3.36), ranging from patches of 0.20 ha to large forest clearings of 229 ha. ...
Preprint
Full-text available
In the Neotropics, the integration of remotely sensed products to understand socioecological processes at local scales is limited by the physical difficulties and financial costs of collecting field data to train and validate these models. In this study, we used carbon density and tree species richness models generated from ALOS-2 PALSAR 2 imagery and national scale forest inventory data and compared these products to a Landsat-based continuous change detection product for the 2000-2021 period. This was performed to evaluate forest clearing dynamics in and around the Puuc Biocultural State Reserve (PBSR) in Mexico. The estimated error-adjusted area of detected annual forest clearings from the year 2000 until the year 2021 was 230,511 ha in total (+19,979 ha). The analysis of annual forest clearing frequency and area suggests that although forest clearing was significantly more intensive outside of the PBSR than within the PBSR during the entire 2000-2021 period, there is no evidence suggesting that the frequency and magnitude of forest clearing has changed over the years after the creation of the PBSR in 2011. The emerging hotspot analysis shows, however, that forest clearing spatiotemporal clustering (hotspots) during the 2012-2021 period was less widespread and mostly confined to areas outside the PBSR. In addition, the analysis shows forest clearing clustering is on a downward trend within the PBSR. After comparing forest clearing events to carbon density and tree species richness models, our data also suggests that land owners within the PBSR might be practicing longer barbecho (fallow) periods in contrast to land owners outside the PBSR allowing forests to attain higher carbon density and tree species richness and hence better soil nutrient recovery after land abandonment. In conclusion, our results show that the PBSR effectively acts as stabilizing forest management scheme that minimizes the impact of productive activities by lowering the frequency of forest clearing events and preserving late secondary forests. We recommend the continuation of efforts for providing alternative optimal field data collection strategies and modeling techniques to spatially predict key tropical forest attributes. The combination of these models with continuous change detection datasets will allow to reveal underlying ecological processes and generate information better adapted to forest governance scales.
... Burned Area product (Giglio et al., 2018) (Griffiths et al., 2018), whereas persistent cloud cover, always a challenge for optical remote sensing, becomes an important constraint in the selection of imagery for forest change detection Mean temperature and annual precipitation series in the BLA at 0.5°s patial resolution are provided by the World Bank Group's portal at https://climateknowledgeportal.worldbank.org/country/brazil/climate-data-historical ...
Thesis
Das Amazonasgebiet hat in den letzten Jahrzehnten eine Intensivierung der menschlichen Aktivitäten erfahren, die in Verbindung mit häufigen schweren Dürren die Umwelt anfälliger für Brände gemacht hat. In dieser Dissertation wurden Fernerkundungsdaten analysiert, um die räumlich-zeitliche Verteilung der Feuer in den letzten 20 Jahren im brasilianischen Amazonasgebiet umfassend zu untersuchen und die verschiedenen Brandursachen zu entschlüsseln. (I) Die erste Forschungsarbeit wertete die Verteilung der verbrannten Fläche aus und zeigte, dass die meisten Brände auf bewirtschafteten Weiden und in den immergrünen Tropenwäldern auftraten, was die Behauptung stützt, dass ihr Auftreten stark auf anthropogene Landnutzungsänderungen reagiert. Die Ergebnisse zeigten auch, dass weder Entwaldung noch Walddegradierung mit Waldbränden korrelierte, wohl aber Feuer, die auf Weiden oder Ackerflächen gelegt wurden und in den angrenzenden Wald übergesprungen sind. (II) Die zweite Forschungsarbeit analysierte einzelne Brände, die durch den auf komplexen Netzwerken basierenden FireTracks-Algorithmus identifiziert wurden. Der Algorithmus wurde verwendet, um Feuerregime für sechs verschiedene Landnutzungsklassen zu ermitteln. Die integrierte Größe, Dauer, Intensität und Ausbreitungsrate dieser räumlich-zeitlichen Brandcluster in den verschiedenen Landnutzungstypen zeigte auf, wie seltene Waldbrände, die natürlicherweise nicht in immergrünen tropischen Wäldern vorkommen, sich zu einem Feuerregime entwickelten, das für Savannenbrände typisch ist. (III) Die dritte Forschungsarbeit analysierte extreme, d. h. die intensivsten Einzelfeuer in immergrünen tropischen Wäldern, und zeigte deren großen Anteil an der insgesamt verbrannten Waldfläche. Während der globale Klimawandel das Potenzial hat, die Trockenheit zu verstärken, sind die anthropogenen Ursachen der Waldzerstörung die Zündquellen, die die Verteilung extremer Brände in den empfindlichen tropischen Wäldern bestimmen.
... However, the main one is ecosystem services, such as forests, which are costless if preserved and maintained by the Brazilian Environmental Law requiring native vegetation protection. Tropical forests are of great importance for climate regulation, carbon sequestration, and high levels of biodiversity (Griffiths et al., 2018;Silva & Cardoso, 2021). They are active carbon sequestration instruments, transforming organic vegetable matter into a litter that enriches the soil (Silva & Cardoso, 2021). ...
Article
Full-text available
Climate change is a growing concern worldwide. Indisputably, continuous anthropogenic emissions of greenhouse gases (GHG) increasingly impact ecosystems, human life, and agricultural production. Strategies for climate change impact mitigation are as necessary as strategies for forest preservation, so these may continue acting as essential ecosystems which capture atmospheric carbon dioxide CO2. This study has a twofold objective. First, to analyze the primary CO2 emissions in Mato Grosso state soybean cultivation, one of the leading products of Brazilian agribusiness chains. Second, to determine the carbon sequestration potential in the different biomes of the states' macro-regions and check if the biomes' forested areas can mitigate GHG by neutralizing soy activity emissions, as mandated by local environmental laws. We collect soybean production data from around Mato Grosso, computing the crop areas, macro-regions, and biomes. The emissions of soybean production were calculated and compared with the carbon sequestration potential of the biomes without agricultural activities. The biomes and forest areas correspond to the preserved area by the Brazilian environmental preservation law. The results indicated that carbon sequestration by the biome areas was higher than CO2 emissions caused by anthropogenic activities during soybean cultivation and that environmental laws sufficed to mitigate such emissions. Soybean farmers must cooperate in adopting low-carbon agricultural practices, preserving the legal areas for future generations.
... Traditional global segmentation is often affected by the interaction between the change points. Therefore, there are more uncertainties for eucalyptus forests that operate in short rotation, and Griffiths, Jakimow, and Hostert (2018) stated that approaches utilizing temporal segmentation (Hermosilla et al. 2016;Kennedy, Yang, and Cohen 2010) do not readily capture fast land-use changes, such as swift postdeforestation dynamics or circular slash-and-burn land-use practices. Our results show that the random localization calculation scheme was very robust and consistent at different lengths of NDVI time series and very short spacing between the breakpoints. ...
Article
Full-text available
Obtaining robust change-detection results and reconstructing planting history are important bases for conducting forest resource monitoring and management. The existence of multiple change points in a very short period can lead to a global segmentation method incorrectly locate the change points, because they could impact each other during model initialization. This is especially true for monitoring plantations such as eucalyptus, which has a unique growth cycle with short rotation periods and frequent disturbances. In this study, we proposed a method to find critical change points in a normalized difference vegetation index (NDVI) time series by combining random localization segmentation and the Chow test. Features of the NDVI time series calculated on the divided segments and change points were used to train a Random Forest classifier for continuous land-cover classification. The proposed method was successfully applied to a eucalyptus plantation for identifying the management history, including harvest time, generation, rotation cycle, and stand age. The results show that our method is robust for different lengths of NDVI time series, and can detect short-interval (cut and stability) change points more accurately than the global segmentation method. The overall accuracy of identification was 80.5%, and successive generations in 2021 were mainly first- and second-generation, accounting for 69.0% and 27.9% of the total eucalyptus area, respectively. The rotation cycle of eucalyptus plantation is usually 5–8 years for 66.9% of the total area. The eucalyptus age was accurately estimated with an R² value of 0.91 and RMSE of 13.3 months. One-year-old eucalyptus plantations accounted for the highest percentage of 14.5%, followed by seven-year-old plantations (12.9%). This study provides an important research basis for accurately monitoring the rotation processes of short-period plantations, assessing their timber yield and conducting carbon- and water-cycle research.
... Increasing access to satellite images along with new processing capabilities offer new opportunities for understanding frontier dynamics at unprecedented temporal and spatial resolution (Gorelick et al 2017, Wulder et al 2019, Woodcock et al 2020, yet these opportunities have so far not been explored. Prior work on assessing frontiers has mostly focused on mapping deforestation (Hansen et al 2013, Müller et al 2016, Griffiths et al 2018, Vancutsem et al 2021, what follows deforestation (Zalles et al 2019, Souza et al 2020, Song et al 2021 or, most recently, who drives deforestation frontiers (Curtis et al 2018. The question of how frontier dynamics unfold, beyond identifying hotspots of deforestation (Harris et al 2017, Potapov et al 2019, remains largely unexplored. ...
Article
Full-text available
Agricultural expansion into tropical and subtropical forests often leads to major social-ecological trade-offs. Yet, despite ever-more detailed information on where deforestation occurs, how agriculture expands into forests remains unclear, which is hampered by a lackof spatially and temporally detailed reconstruction of agricultural expansion. Here, we developed and mapped a novel set of metrics that quantify agricultural frontier processes at unprecedented spatial and temporal detail. Specifically, we first derived consistent annual time series of land-use/cover to, second, describe archetypical patterns of frontier expansion, pertaining to speed, diffusion and activity of deforestation, as well as post-deforestation land use. We exemplify this approach for understanding agricultural frontier expansion across the entire South American Chaco, a global deforestation hotspot. Our study provides three major insights. First, agricultural expansion has been rampant in the Chaco, with more than 19.3 million ha of woodlands converted between 1985 and 2020, including a surge in deforestation after 2019. Second, land-use trajectories connected to frontier processes have changed in major ways over the 35-year study period we studied, including substantial regional variations. For instance, while ranching expansion drove most of the deforestation in the 1980s and 1990s, cropland expansion dominated during the mid-2000s in Argentina, but not in Paraguay. Similarly, 40% of all areas deforested were initially used for ranching, but later on converted to cropping. Accounting for post-deforestation land-use change is thus needed to properly attribute deforestation and associated environmental impacts, such as carbon emissions or biodiversity loss, to commodities. Finally, we identified major, recurrent frontier types that may be a useful spatial template for land governance to match policies to specific frontier situations. Collectively, our study reveals the diversity of frontier processes and how frontier metrics can capture and structure this diversity to uncover major patterns of human-nature interactions, which can be used to guide spatially-targeted policies.
... This is analogous to postclassification cleaning of hard classification change detections, by removing illogical transitions (e.g. Griffiths et al., 2018) or applying statistical techniques such as Hidden Markov Models (e.g. Abercrombie & Friedl, 2016). ...
Article
Full-text available
The Ngorongoro Conservation Area (NCA) of Tanzania, is globally significant for biodiversity conservation due to the presence of iconic fauna, and, since 1959 has been managed as a unique multiple land‐use areas to mutually benefit wildlife and indigenous residents. Understating vegetation dynamics and ongoing land cover change processes in protected areas is important to protect biodiversity and ensure sustainable development. However, land cover changes in savannahs are especially difficult, as changes are often long‐term and subtle. Here, we demonstrate a Landsat‐based monitoring strategy incorporating (i) regression‐based unmixing for the accurate mapping of the fraction of the different land cover types, and (ii) a combination of linear regression and the BFAST trend break analysis technique for mapping and quantifying land cover changes. Using Google Earth Pro and the EnMap‐Box software, the fractional cover of the main land cover types of the NCA were accurately mapped for the first time, namely bareland, bushland, cropland, forest, grassland, montane heath, shrubland, water and woodland. Our results show that the main changes occurring in the NCA are the degradation of upland forests into bushland: we exemplify this with a case study in the Lerai Forest; and found declines in grassland and co‐incident increases in shrubland in the Serengeti Plains, suggesting woody encroachment. These changes threaten the wellbeing of livestock, the livelihoods of resident pastoralists and of the wildlife dependent on these grazing areas. Some of the land cover changes may be occurring naturally and caused by herbivory, rainfall patterns and vegetation succession, but many are linked to human activity, specifically, management policies, tourism development and the increase in human population and livestock. Our study provides for the first time much needed and highly accurate information on long‐term land cover changes in the NCA that can support the sustainable management and conservation of this unique UNESCO World Heritage Site. The Ngorongoro Conservation Area (NCA), a UNESCO World Heritage Site, is globally important for biodiversity conservation due to the presence of iconic megafauna. For decades now, the NCA is experiencing a number of notoriously difficult to address challenges; understanding its land cover dynamics is therefore increasingly important to improve habitat monitoring, preserve biodiversity and ensure sustainable development. We used multi‐temporal Landsat data spanning 35 years and a combination of regression‐based unmixing, linear regression and a trend break analysis to map and quantify the land cover dynamics in the area. We found a decrease in forest and grassland cover as well as a significant amount of woody encroachment which is often linked to land degradation in African savannahs. These changes are consistent with other savannah ecosystems and pose a threat to the wellbeing of livestock, the livelihoods of the pastoralist communities, and the wildlife of the NCA.
... Increasing access to satellite images along with new processing capabilities offer new opportunities for understanding frontier dynamics at unprecedented temporal and spatial resolution (Gorelick et al 2017, Wulder et al 2019, Woodcock et al 2020, yet these opportunities have so far not been explored. Prior work on assessing frontiers has mostly focused on mapping deforestation (Hansen et al 2013, Müller et al 2016, Griffiths et al 2018, Vancutsem et al 2021, what follows deforestation (Zalles et al 2019, Souza et al 2020, Song et al 2021 or, most recently, who drives deforestation frontiers (Curtis et al 2018. The question of how frontier dynamics unfold, beyond identifying hotspots of deforestation (Harris et al 2017, Potapov et al 2019, remains largely unexplored. ...
Preprint
Agricultural expansion into tropical and subtropical forests often leads to major social-ecological trade-offs. Yet, despite ever-more detailed information on where deforestation occurs, how agriculture expands into forests remains unclear. Here, we developed and mapped a novel set of metrics that quantify agricultural frontier processes at unprecedented spatial and temporal detail. Specifically, we first derived consistent time series of land-use/cover to, second, describe archetypical patterns of frontier expansion, pertaining to the speed, the diffusion and activity of deforestation, as well as post-deforestation land use. We exemplify this approach for understanding agricultural frontier expansion across the entire South American Chaco (1.1 million km2), a global deforestation hotspot. Our study provides three major insights. First, agricultural expansion has been rampant in the Chaco, with more than 19.3 million ha of woodlands converted between 1985 and 2020, including a surge in deforestation after 2019. Second, land-use trajectories connected to frontier processes have changed in major ways over the 35-year study period we studied. For instance, while ranching expansion drove most of the deforestation in the 1980s and 1990s, cropland expansion dominated during the mid-2000s in Argentina, but not in Paraguay. Similarly, 40% of all areas deforested were initially used for ranching, but later on converted to cropping. Accounting for post-deforestation land-use change is thus needed to properly attribute deforestation and associated environmental impacts, such as carbon emissions or biodiversity loss, to commodities. Finally, we identified major, recurrent frontier types that may be a useful spatial template for land governance to match policies to specific frontier situations. Collectively, our study reveals the diversity of frontier processes and how frontier metrics can capture and structure this diversity for guiding spatially targeted policies, and for uncovering high-level patterns of human-nature interactions.
... A key difference between regional and global products is that at regional scales, validation according to best practices (Olofsson et al., 2014) is routinely done and expected (Griffiths et al., 2018;Tulbure et al., 2016). At the global scale, however, this is much more challenging, and not many maps are validated according to best practices, which precludes their usage, with notable exceptions (Pickens et al., 2020). ...
Article
Full-text available
Unprecedented amounts of analysis‐ready Earth Observation (EO) data, combined with increasing computational power and new algorithms, offer novel opportunities for analysing ecosystem dynamics across large geographic extents, and to support conservation planning and action. Much research effort has gone into developing global EO‐based land‐cover and land‐use datasets, including tree cover, crop types, and surface water dynamics. Yet there are inherent trade‐offs between regional and global EO products pertaining to class legends, availability of training/validation data, and accuracy. Acknowledging and understanding these trade‐offs is paramount for both developing EO products and for answering science questions relevant for ecology or conservation studies based on these data. Here we provide context on the development of global EO‐based land‐cover and land‐use datasets, and outline advantages and disadvantages of both regional and global datasets. We argue that both types of EO‐derived land‐cover datasets can be preferable, with regional data providing the context‐specificity that is often required for policy making and implementation (e.g., land‐use and management, conservation planning, payment schemes for ecosystem services), making use of regional knowledge, particularly important when moving from land cover to actors. Ensuring that global and regional land‐cover and land‐use products derived based on EO data are compatible and nested, both in terms of class legends and accuracy assessment, should be a key consideration when developing such data. Open access high‐quality training and validation data derived as part of such efforts are of utmost importance. Likewise, global efforts to generate sets of essential variables for climate change, biodiversity, or eventually land use, which often require land‐cover maps as inputs, should consider regionalized, hierarchical approaches to not sacrifice regional context. Global change impacts manifest in regions, and so must the policy and planning responses to these challenges. EO data should embrace that regions matter, perhaps more than ever, in an age of global data availability and processing.
... The indicators of deforestation areas refer to areas deforested in 2019, according to data from the National Institute for Space Research (INPE). These areas were not necessarily deforested for soybean production, but in principle for logging, followed by livestock raising, and then into agricultural areas [38,44,45]. The remaining areas are the existing forests of each biome until the year 2019 [38]. ...
Article
Full-text available
Citation: Toloi, M.N.V.; Bonilla, S.H.; Toloi, R.C.; Silva, H.R.O.; Nääs, I.d.A. Development Indicators and Soybean Production in Brazil. Agriculture 2021, 11, 1164. https://doi.
... Remote sensing (Chen, Chuvieco et al. 2021) is an effective tool for Earth Observation (EO) and can be used to detect spatial variations across large regions on multiple scales. In literature, it is commonly used to evaluate environmental changes, such as atmosphere pollution (Wu, Zhang et al. 2020), deforestation (Griffiths, Jakimow et al. 2018), and urban development on scales ranging from countrywide to the worldwide. In general, two strategies were commonly used to estimate the environmental degradation by remote sensing datasets. ...
Article
Full-text available
The China-Europe Railway Express (CER-Express), not only promoted the cooperation between countries and regions across Asia and Europe but also gave rise to remarkable changes in landcover and had a profound effect on the natural environment along the railway in recent years. Effective ways to monitor and assess ecological changes are urgently needed to ensure sustainable development of CER-Express. There are very few existing environmental monitoring studies focusing on the area along the CERExpress. In this paper, we present a study of environmental degradation, which occurred during the construction and operation of CER-Express from 2010 to 2018, based on a comprehensive evaluation index (CEI), which takes three environmental indicators into account and provides a timely and reliable evaluation of environmental changes at large scales. In addition, the environment conditions of the regions and countries along the CER-Express have been quantified and comparatively studied at different scales over different periods, using histograms of mean CEI values. Furthermore, specific causes of environmental degradation in the rail-intensive countries and small-area countries along the railway are discussed. Our results show that the environmental degradation can be detected in most of the rail-intensive countries, such as Germany, Poland, Austria, and Czech Republic along the railway. Therefore, to ensure sustainability of the CER-Express, environmental protection along the railway should be paid more attention to and a reasonable arrangement for the exploitation of CER-Express devised.
... We used an extensive database of training samples [30] and hand-digitized deforestation polygons between 2014 and 2016 from GUYRA Paraguay http://guyra. org.py/informe-deforestacion. We used these training data to parameterize a time-calibrated random forest classifier and classified 31 annual woodland loss maps between 1985 and 2016 [47]. Each map used satellite data from that year and the previous year, to ensure consistency between years. ...
Article
Full-text available
Land-use change is a root cause of the extinction crisis, but links between habitat change and biodiversity loss are not fully understood. While there is evidence that habitat loss is an important extinction driver, the relevance of habitat fragmentation remains debated. Moreover, while time delays of biodiversity responses to habitat transformation are well-documented, time-delayed effects have been ignored in the habitat loss versus fragmentation debate. Here, using a hierarchical Bayesian multi-species occupancy framework, we systematically tested for time-delayed responses of bird and mammal communities to habitat loss and to habitat fragmentation. We focused on the Argentine Chaco, where deforestation has been widespread recently. We used an extensive field dataset on birds and mammals, along with a time series of annual woodland maps from 1985 to 2016 covering recent and historical habitat transformations. Contemporary habitat amount explained bird and mammal occupancy better than past habitat amount. However, occupancy was affected more by the past rather than recent fragmentation, indicating a time-delayed response to fragmentation. Considering past landscape patterns is therefore crucial for understanding current biodiversity patterns. Not accounting for land-use history ignores the possibility of extinction debt and can thus obscure impacts of fragmentation, potentially explaining contrasting findings of habitat loss versus fragmentation studies.
... 2015 (current situation, after a partial revival of the agricultural sector). Image composites are gap-and cloud-free mosaics based on Landsat images (Griffiths, Jakimow, & Hostert, 2018). For each of the three time steps, we calculated three composites centred on spring (Julian day 121), summer (day 180) and fall (day 260) to capture phenology differences that are important for mapping cropland-grassland dynamics (Baumann et al., 2011). ...
Article
Full-text available
Large and ecologically functioning steppe complexes have been lost historically across the globe, but recent land‐use changes may allow the reversal of this trend in some regions. We aimed to develop and map indicators of changing human influence using satellite imagery and historical maps, and to use these indicators to identify areas for broad‐scale steppe rewilding.
... In particular, an area was chosen that is located between Araweté Igarapé Ipixuna and Kayapó Indigenous Territory (see Figure 1), which covers an area of 48,838 km 2 . The final decision to perform the research on this area was supported by the evident variation of the rainforest highlighted in the academic literature [37,38]. The region close to São Félix do Xingu, the main city in the AOI, appears as a deforestation hot spot as it has suffered from pasture expansion, cattle ranching, road construction, land occupations and agrarian development [39][40][41]. ...
Article
Full-text available
Deforestation causes diverse and profound consequences for the environment and species. Direct or indirect effects can be related to climate change, biodiversity loss, soil erosion, floods, landslides, etc. As such a significant process, timely and continuous monitoring of forest dynamics is important, to constantly follow existing policies and develop new mitigation measures. The present work had the aim of mapping and monitoring the forest change from 2000 to 2019 and of simulating the future forest development of a rainforest region located in the Pará state, Brazil. The land cover dynamics were mapped at five-year intervals based on a supervised classification model deployed on the cloud processing platform Google Earth Engine. Besides the benefits of reduced computational time, the service is coupled with a vast data catalogue providing useful access to global products, such as multispectral images of the missions Landsat five, seven, eight and Sentinel-2. The validation procedures were done through photointerpretation of high-resolution panchromatic images obtained from CBERS (China-Brazil Earth Resources Satellite). The more than satisfactory results allowed an estimation of peak deforestation rates for the period 2000-2006; for the period 2006-2015, a significant decrease and stabilization, followed by a slight increase till 2019. Based on the derived trends a forest dynamics was simulated for the period 2019-2028, estimating a decrease in the deforestation rate. These results demonstrate that such a fusion of satellite observations, machine learning, and cloud processing, benefits the analysis of the forest dynamics and can provide useful information for the development of forest policies.
... This is the first time that LULC change has been quantified in all Brazilian biomes with this degree of spatial detail (i.e., at 30 m pixel size) using +30-year time-series Landsat data. Until now, this LULC change information in Brazil was either restricted in space and time, covering a few biomes and short periods of time (e.g., [53][54][55]), or long time-series, but focusing on deforestation in the portion of one of the biomes [56]. Coarser spatial resolution remote sensing images have also been used to map LULC using Google Earth Engine covering all biomes in a single year [57], and global LULC products [31] are available with limited inputs from local experts. ...
Article
Full-text available
Brazil has a monitoring system to track annual forest conversion in the Amazon and most recently to monitor the Cerrado biome. However, there is still a gap of annual land use and land cover (LULC) information in all Brazilian biomes in the country. Existing countrywide efforts to map land use and land cover lack regularly updates and high spatial resolution time-series data to better understand historical land use and land cover dynamics, and the subsequent impacts in the country biomes. In this study, we described a novel approach and the results achieved by a multi-disciplinary network called MapBiomas to reconstruct annual land use and land cover information between 1985 and 2017 for Brazil, based on random forest applied to Landsat archive using Google Earth Engine. We mapped five major classes: forest, non-forest natural formation, farming, non-vegetated areas, and water. These classes were broken into two sub-classification levels leading to the most comprehensive and detailed mapping for the country at a 30 m pixel resolution. The average overall accuracy of the land use and land cover time-series, based on a stratified random sample of 75,000 pixel locations, was 89% ranging from 73 to 95% in the biomes. The 33 years of LULC change data series revealed that Brazil lost 71 Mha of natural vegetation, mostly to cattle ranching and agriculture activities. Pasture expanded by 46% from 1985 to 2017, and agriculture by 172%, mostly replacing old pasture fields. We also identified that 86 Mha of the converted native vegetation was undergoing some level of regrowth. Several applications of the MapBiomas dataset are underway, suggesting that reconstructing historical land use and land cover change maps is useful for advancing the science and to guide social, economic and environmental policy decision-making processes in Brazil.
... 2015 (current situation, after a partial revival of the agricultural sector). Image composites are gap-and cloud-free mosaics based on Landsat images (Griffiths, Jakimow, & Hostert, 2018). For each of the three time steps, we calculated three composites centred on spring (Julian day 121), summer (day 180) and fall (day 260) to capture phenology differences that are important for mapping cropland-grassland dynamics (Baumann et al., 2011). ...
Article
Full-text available
Large and ecologically functioning steppe complexes have been lost historically across the globe, but recent land‐use changes may allow the reversal of this trend in some regions. We aimed to develop and map indicators of changing human influence using satellite imagery and historical maps, and to use these indicators to identify areas for broad‐scale steppe rewilding. We mapped decreasing human influence indicated by cropland abandonment, declining grazing pressure and rural outmigration in the steppes of northern Kazakhstan. We did this by processing ~5,500 Landsat scenes to map changes in cropland between 1990 and 2015, and by digitizing Soviet topographic maps and examining recent high‐resolution satellite imagery to assess the degree of abandonment of >2,000 settlements and >1,300 livestock stations. We combined this information into a human influence index (HI), mapped changes in HI to highlight where rewilding might take place and assessed how this affected the connectivity of steppe habitat. Across our study area, about 6.2 million ha of cropland were abandoned (30.5%), 14% of all settlements were fully and 81% partly abandoned, and 76% of livestock stations were completely dismantled between 1990 and 2015, suggesting substantially decreasing human pressure across vast areas. This resulted in increased connectivity of steppe habitat. The steppes of Eurasia are experiencing massively declining human influence, suggesting large‐scale passive rewilding is taking place. Many of these areas are now important for the connectivity of the wider steppe landscape and can provide habitat for endangered megafauna such as the critically endangered saiga antelope. Yet, this window of opportunity may soon close, as recultivation of abandoned cropland is gaining momentum. Our aggregate human influence index captures key components of rewilding and can help to devise strategies for fostering large, connected networks of protected areas in the steppe.
... 2015 (current situation, after a partial revival of the agricultural sector). Image composites are gap-and cloud-free mosaics based on Landsat images (Griffiths et al. 2014;Griffiths et al. 2018). ...
Thesis
Full-text available
Temperate grasslands are widespread, provide important ecosystem services, and often offer good conditions for agriculture. As a result, many temperate grasslands are undergoing agricultural land-use change. While in most world regions these changes result in expansion and intensification of agriculture, some regions exhibit the opposite trajectory, providing opportunities for balancing trade-offs between food production and grassland restoration. Land abandonment may lead to negative ecological consequences, though, such as increasing fire frequency or severity. The temperate steppes of Kazakhstan are one of the world regions that experienced massive changes in land management intensity and widespread land-use change after the breakdown of the Soviet Union. Cropping and grazing regime changes across the steppes of Kazakhstan are understudied, and related spatio-temporal changes, e.g. in fire regimes, are still poorly understood. The main research goal of this thesis was accordingly to develop a methodology to map related change at appropriate scales and to provide novel datasets with high spatial and temporal detail to enhance our understanding of how the coupled human-environment in Northern Kazakhstan has changed since the 1980s. An approach was developed to identify the timing of post-Soviet cropland abandonment and recultivation in northern Kazakhstan based on annual Landsat time series. Knowing the timing of abandonment allowed for deeper insights into what drives these dynamics: for example, recultivation after 2007 happened mainly on land that had been abandoned latest. Likewise, knowing the timing of abandonment allowed for substantially more precise estimates of soil organic carbon sequestration. Mapping changes in fire regimes (i.e. extent, number and size of fires) highlighted a sevenfold increase in burnt area and an eightfold increase in number of fires after the breakdown of the Soviet Union. Agricultural burning as well as cropland and pasture abandonment were associated with increased fire risk. It was therefore important to provide better estimates on how grazing pressure changed after the dissolution of the Soviet Union. Grazing probabilities, derived from a number of spectral indices using a random forest, were found to provide the best metrics to capture grazing pressure. The analysis revealed a general decline in grazing pressure in the Kazakh steppe after 1992. The effect was mostly pronounced near abandoned livestock stations, and significantly increased with distance from such points. Collectively, the analyses in this dissertation highlight how dense records of Landsat images can be utilized to better understand land use changes and the ecology of steppes across large areas. The datasets developed within this thesis specifically allow to disentangle the processes leading to and the impacts of agricultural abandonment in the temperate Kazakh steppes, and may potentially be used to support decision-making in land-use and conservation planning.
... Increasing availability of remote sensing and social sensing data as well as methodological advances within digital image processing and data fusion is opening windows of opportunity for improved analyses of land use dynamics. Such technological advances are also expected to lead to a better understanding of drivers and pressures of land use changes [38,39] and impacts of human activities [40][41][42]. ...
Article
Changes of land systems are largely a consequence of human decision making at multiple scales, from local land management to national land use planning and global trade agreements. Improved understanding of the status, trend, and consequences of land changes is relevant to the engagement of Land System Science (LSS) in transformations for sustainability. While remote sensing has been a major tool for documenting land cover change, new opportunities are emerging to understand land as a socio-ecological system by integrating crowd-sourced social sensing data (e.g. through mobile technologies). Increased access to remote sensing time series and social sensing data are improving mapping of land and providing new information on the coupled social-ecological system for better land governance. Given advances in data availability and machine learning algorithms, land mapping efforts have evolved to more sophisticated analysis strategies. In addition, the emergence of planetary-scale geospatial analysis platforms facilitates regional and global land change monitoring in a rapid, scalable, and convenient way. This review aims to document current status and prospects on data, algorithms, and processing platforms in the era of ‘Big Earth Data’, providing examples of how it can contribute to LSS at the interface of normative and policy concerns, including those in support of internationally agreed environmental goals and land governance.
... Data from TM and ETM+ have been combined via compositing approaches that use data corrected to surface reflectance (Flood 2013;Griffiths et al. 2013) via the Landsat Ecosystem Disturbance Adaptive Processing System (LEDAPS) (Masek et al. 2006;Feng et al. 2013). More recently, OLI data has been integrated into these compositing approaches Griffiths et al. 2018). The United States Geological Survey (USGS) now provides Landsat surface reflectance products as a Level-2 Science Product, with data from Landsats 4-7 corrected with LEDAPS (USGS 2012; 2018a), whereas OLI is corrected with Landsat Surface Reflectance Code (LaSRC; USGS 2018b, Vermote et al. 2016). ...
Thesis
Full-text available
Information needs associated with forest monitoring have become increasingly complex. Data to support these information needs are required to be systematically generated, spatially exhaustive, spatially explicit, and to capture changes at a spatial and temporal resolution that is commensurate with both natural and anthropogenic impacts. Moreover, reporting obligations impose additional expectations of transparency, repeatability, and data provenance. The overall objective of this dissertation was to address these needs and improve capacity for large-area monitoring of forest disturbance and subsequent recovery. Landsat time series (LTS) enhance opportunities for forest monitoring, particularly for post-disturbance recovery assessments, while best-available pixel (BAP) compositing approaches allow LTS approaches to be applied over large forest extents. In substudies I and IV, forest monitoring information needs were identified and linked to image compositing criteria and data availability in Canada and Finland. In substudy II, methods were developed and demonstrated for generating large-area, gap-filled Landsat BAP image composites that preserve detected changes, generate continuous change metrics, and provide foundational, annual data to support forest monitoring. In substudy III a national monitoring framework was prototyped at scale over the 650 Mha of Canada’s forest ecosystems, providing a detailed analysis of areas disturbed by wildfire and harvest for a 25-year period (1985–2010), as well as characterizing short- and long-term recovery. New insights on spectral recovery metrics were provided by substudies V and VI. In substudies V, the utility of spectral measures of recovery were evaluated and confirmed against benchmarks of forest cover and height derived from airborne laser scanning data. In substudy VI the influence of field-measured structure and composition on spectral recovery were examined and quantified. By focusing on four key aspects of forest monitoring systems: information needs, data availability, methods development, and information outcomes, the component studies demonstrated that combining BAP compositing and LTS analysis approaches provides data with the requisite characteristics to support large-area forest monitoring, while also enabling a more comprehensive assessment of forest disturbance and recovery.
... Mudanças de padrão do uso do solo são alguns dos fatores que ocasionam alterações no clima na região da Amazônia (SPRACKLEN; GARCIA-CARRERAS, 2015;ZEMP et al., 2017). O desmatamento e as queimadas, por exemplo, são questões presentes que impactam diretamente no clima tropical (AYALA et al., 2016), tais como o aumento da temperatura e redução de chuvas, causando uma maior sensibilidade nos ecossistemas (SEDDON et al., 2016;GRIFFITHS et al., 2018). ...
Conference Paper
Full-text available
Mudanças do uso da terra na Amazônia tornaram-se mais frequentes nas últimas décadas, ocasionando intensas alterações ambientais e climáticas. O objetivo geral do estudo é analisar os padrões espaciais do desmatamento e associar a influência das ocorrências de queimadas com a variação anual da temperatura na APA Triunfo do Xingu, no período de 2005 a 2015. Utilizou-se dois produtos do Instituto Nacional de Pesquisas Espaciais (INPE) denominado PRODES e BDQueimadas, e dados meteorológicos de temperatura do ar obtidos da interpolação de estações pluviométricas e meteorológicas desenvolvida pelo CLIMA. Os resultados indicam que a APA Triunfo do Xingu possui uma área territorial muito degradada, se comparada a outras unidades de conservação e a intensidade dos focos de calor na região não é um efeito antrópico contínuo. Além disso, considerando a influência do número de queimadas com as variáveis meteorológicas, os maiores picos de focos de calor estão fortemente associados com as elevadas temperaturas, explicando a variabilidade anual das condições de tempo vigente acima do normal. Adicionalmente, o número de focos de calor e a variação da temperatura do ar caracterizou-se por uma linha decrescente durante o período de existência das fases do Plano de Ação para Prevenção e Controle do Desmatamento na Amazônia (PPCDAm), sendo essas duas variáveis impulsionadas nos últimos anos pela ocorrência de mecanismo de precipitação de El Niño.
Article
Full-text available
The Landsat archive is one of the richest Earth observation datasets available and provides long-term data at fairly high temporal and spatial resolution globally. Temporal aggregation is frequently used to condense single observations into a more digestible feature space that provides spatially gap-free data to fulfill demands of many processing strategies that rely on homogeneous data coverage across a large area, e.g., machine learning-based land cover classification. Spectral Temporal Metrics (STMs) represent a conceptually simple feature space wherein multiple observations are temporally aggregated by statistical functions. The quality and inter-annual consistency of STMs is affected by data availability, including usable clear-sky observations that vary in time and space due to satellite lifecycles, sensor failures, changes of observation modes, climate regimes, orbital overlaps, as well as inter-annual variability of cloud cover. However, the relationship between data availability and STM consistency between years is still poorly understood, especially as differences in STMs between years can both result from inter-annual variability in data availability, as well as inter-annual variability of land surfaces. In this study, we systematically quantify the effect of inter-annual data availability on annual STMs for the years 1984–2019, while completely controlling for inter-annual land surface changes. Our results are expected to help assess where on Earth, and in what periods, specific metrics can be used or should be avoided when multi-annual consistency is required. We synthesized a global, nearly gap-free reference time series at daily temporal resolution from MODIS data. This “baseline” was subsequently degraded with actual annual Landsat mission observation scenarios resulting in synthetic annual time series that only differ with respect to data availability. Based on the differences between STMs generated from the baseline, and STMs computed from the degraded time series, we statistically quantified the accuracy, precision, and uncertainty (APU) for various STMs across the Landsat spectral bands. We compared the performance against a reasonable specification, i.e., a tolerated error. We aggregated APU metrics along climate zones annually to carve out regional and temporal differences. We found that huge regional differences exist, with the highest quality and consistency in arid climates (i.e., APU within specification). Errors in fully humid snow climates are high, yet systematic (biased but repeatable), whereas equatorial and temperate climates are characterized by unbiased but uncertain metrics, where accuracy or precision and uncertainty can exceed specification by a factor of three or more. Quality generally increased with time as a response to improved observation modes and data storage commitment, e.g., uncertainty improved from one sensor availability period to the next in >90% of all climate zones for the near infrared average – with the exception of the Landsat 7 scanline corrector failure in 2003 where quality decreased again in 62% of climate zones. We also derived and tested different measures of STM quality and found that the seasonal distribution of clear-sky observations is more important than the number of observations, e.g., the near infrared standard deviation's accuracy can be explained with an R2 of 0.55, and 0.78 by the number of observations, and maximum time between subsequent observations in Cfb climates, respectively. Furthermore, our findings revealed how many observations, or how short the largest gap between consecutive observations must be to still produce reliable metrics (e.g., a maximum gaps of 42–45 days to obtain tolerated uncertainty of the near infrared average and standard deviation in Cfb climates), which has substantial implications for the design of downstream applications relying on multi-annual STM. This study provides the tools for a global and systematic assessment of inter-annual STM consistency while controlling for land-surface dynamics and thereby paves the way for a systematic error quantification in Level 3 products.
Article
Carbon offsets from voluntary avoided-deforestation projects are generated on the basis of performance in relation to ex ante deforestation baselines. We examined the effects of 26 such project sites in six countries on three continents using synthetic control methods for causal inference. We found that most projects have not significantly reduced deforestation. For projects that did, reductions were substantially lower than claimed. This reflects differences between the project ex ante baselines and ex post counterfactuals according to observed deforestation in control areas. Methodologies used to construct deforestation baselines for carbon offset interventions need urgent revisions to correctly attribute reduced deforestation to the projects, thus maintaining both incentives for forest conservation and the integrity of global carbon accounting.
Article
Full-text available
This study aims to determine the management of digital archives at the Syar'iyah Court Office of East Aceh Regency, Indonesia, as well as finding obstacles encountered in their implementation. This research is a qualitative analysis where the subjects of this research were officers who managed digital archives at the Syar'iyah Court office. The data collection techniques employed were interviews, observations, and documentation. The results of this study indicate that the management of digital archives at the Syar'iyah Court Office has not been implemented optimally due to three conditions, namely: 1) The creation of digital archives is hampered due to frequent power outages in East Aceh and Sarana districts which are less supportive, and lack of resources. employees who understand digital archiving issues; 2) The process of borrowing archives carried out at the Syar'iyah Court Office has not been going well because it has not used archive lending procedures such as requesting archives, searching, retrieval of archives, recording, controlling, and storing again, to prevent loss of records; 3) Archive rediscovery still takes quite a long time, ranging from 20 to 30 minutes. The management of digital archives at the Office of the Syar'iyah Idi Court of East Aceh Regency should be improved by proposing additional archiving facilities, namely the latest model scan tool and additional employees who handle digital archive issues.
Article
Full-text available
Water withdrawals for irrigated crop production constitute the largest source of freshwater consumption on Earth. Monitoring the dynamics of irrigated crop cultivation is crucial for tracking crop water consumption, particularly in water-scarce areas. We analyzed changes in water-dependent crop cultivation for 650 000 km ² of Central Asian drylands, including the entire basin of the Amu Darya river, once the largest tributary to the Aral Sea before large-scale irrigation projects grossly reduced the amount of water reaching the river delta. We used Landsat time series to map overall cropland extent, dry season cropping, and cropping frequency in irrigated croplands annually from 1987 to 2019. We scrutinized the emblematic change processes of six localities to discern the underlying causes of these changes. Our unbiased area estimates reveal that between 1988 and 2019, irrigated dry season cropping declined by 1.34 million hectares (Mha), while wet season and double cropping increased by 0.64 Mha and 0.83 Mha, respectively. These results show that the overall extent of cropland in the region remained stable, while higher cropping frequency increased harvested area. The observed changes’ overall effect on water resource use remains elusive: Following the collapse of the Soviet Union, declining dry season cultivation reduced crop water demand while, more recently, increasing cropping frequency raised water consumption. Our analysis provides the first fine-scale analysis of post-Soviet changes in cropping practices of the irrigated areas of Central Asia. Our maps are openly available and can support future assessments of land-system trajectories and, coupled with evapotranspiration estimates, changes in crop water consumption.
Article
Tropical deforestation continues at alarming rates with profound impacts on ecosystems, climate, and livelihoods, prompting renewed commitments to halt its continuation. Although it is well established that agriculture is a dominant driver of deforestation, rates and mechanisms remain disputed and often lack a clear evidence base. We synthesize the best available pantropical evidence to provide clarity on how agriculture drives deforestation. Although most (90 to 99%) deforestation across the tropics 2011 to 2015 was driven by agriculture, only 45 to 65% of deforested land became productive agriculture within a few years. Therefore, ending deforestation likely requires combining measures to create deforestation-free supply chains with landscape governance interventions. We highlight key remaining evidence gaps including deforestation trends, commodity-specific land-use dynamics, and data from tropical dry forests and forests across Africa.
Article
Full-text available
The conversion from primary forest to agriculture drives widespread changes that have the potential to modify the hydroclimatology of the Xingu River Basin. Moreover, climate impacts over eastern Amazonia have been strongly related to pasture and soybean expansion. This study carries out a remote-sensing, spatial-temporal approach to analyze inter-and intra-annual patterns in evapotranspiration (ET) and precipitation (PPT) over pasture and soybean areas in the Xingu River Basin during a 13-year period. We used ET estimates from the Moderate Resolution Imaging Spectroradiometer (MODIS) and PPT estimates from the Tropical Rainfall Measurement Mission (TRMM) satellite. Our results showed that the annual average ET in the pasture was~20% lower than the annual average in soybean areas. We show that PPT is notably higher in the northern part of the Xingu River Basin than the drier southern part. ET, on the other hand, appears to be strongly linked to land-use and land-cover (LULC) patterns in the Xingu River Basin. Lower annual ET averages occur in southern areas where dominant LULC is savanna, pasture, and soybean, while more intense ET is observed over primary forests (northern portion of the basin). The primary finding of our study is related to the fact that the seasonality patterns of ET can be strongly linked to LULC in the Xingu River Basin. Further studies should focus on the relationship between ET, gross primary productivity, and water-use efficiency in order to better understand the coupling between water and carbon cycling over this expanding Amazonian agricultural frontier.
Article
This study used the Landsat 8 OLI satellite image and the supervised classification method to estimate uneven-aged forest stand parameters and land use/cover. The spatial success of classification was also investigated. The overall success rates and Kappa values of the classification were, respectively, 74.7% and 0.75 for actual structural type, 84.6% and 0.80 for crown closure, and 88.35% and 0.81 for land use class, whereas the spatial success of classification on the forest cover type map was 36.91% for actual structural type, 64.74% for crown closure, and 41.78% for land use/cover class. The results revealed that the Landsat 8 OLI image can be used to identify stand parameters and land use/cover class. However, because the spatial success rates were below 50% for the actual structural type and land use/cover class of the stand types, it is not suitable for use in spatial classification determination for these classes.
Article
Multi-spectral spaceborne sensors with different spatial resolutions produce Earth observation (EO) time series (TS) with global coverage. The interactive visualization and interpretation of TS is essential to better understand changes in land-use and land-cover and to extract reference information for model calibration and validation. However, available software tools are often limited to specific sensors or optimized for application-specific visualizations. To overcome these limitations, we developed the EO Time Series Viewer, a free and open source QGIS plugin for user-friendly visualization, interpretation and labeling of multi-sensor TS data. The EO Time Series Viewer (i) combines advantages of spatial, spectral and temporal data visualization concepts that are so far not available in a single tool, (ii) provides maximum flexibility in terms of supported data formats, (iii) minimizes the user-interactions required to load and visualize multi-sensor TS data and (iv) speeds-up labeling of TS data based on enhanced GIS vector tools and formats.
Article
Full-text available
The mapping and monitoring of forest ecosystems on a national scale is key to their management and conservation. Native forests in Uruguay are considered given their importance for biodiversity and ecosystem services. Here we evaluate the spatial distribution of the land cover class ‘hillside and ravine forest’ -a subclass of native forest characterized by patches and transition zones with native grasslands- using Landsat images (30 x 30 m) from 2014 and 2015 and high-resolution images from Google Earth. To evaluate spatial heterogeneity within hillside forests, we then used high-resolution spot images of 1 km2 from 1998-2012 to evaluate differences in the normalized difference vegetation index (ndvi) among canopy coverage categories. The hillside and ravine forest class were characterized as a composite cover class with an average canopy coverage of 69 ± 23%, variability of wich was reflected in ndvi values. The total area of this class in 2015 was estimated as 334,480 ha, somewhat less than an earlier 2008 estimate (384,240 ha). Among the potential errors in delineating hillside forests using Landsat images, there was the classification of «forest» in areas characterized by grassland and a tree canopy cover <25 %. This potential error in delimitation at broader scales led to the overestimation of hillside and ravine forest area in southeastern Uruguay, but an underestimation in northern Uruguay. Our study highlights the large discrepancies in the estimation of the distribution of hillside and ravine forest at different spatial scales, and also indicates the potential of ndvi to evaluate the heterogeneity of this forest within the same cover class.
Thesis
Die tropischen Ökosysteme sind zunehmend steigenden Risiken durch Landnutzungsdruck ausgesetzt. Für die Quantifizierung und Bewertung der ökologischen Vulnerabilität dieser Ökosysteme fehlen allgemeingültige Konzepte und praktisch anwendbare Modelle. Zudem sind die tropischen Waldökosysteme Afrikas wenig erforscht. Im Rahmen dieser Arbeit erfolgt eine konzeptionelle Entwicklung eines räumlich hochauflösenden, multifaktoriellen Landschaftsvulnerabilitätsmodells als Ausdruck für die ökologische Vulnerabilität tropischer Ökosysteme. Das Modell der Landschaftsvulnerabilität (LV = Anfälligkeit der Landschaft für anthropogene Gefährdungen) wird am Fallbeispiel des tropischen Inselökosystems von São Tomé umgesetzt. Die international kaum bekannte Insel São Tomé (859 km²) liegt im Atlantik vor der Westküste des tropischen Zentralafrikas. Aufgrund des Status als Hotspot der Biodiversität mit vielen endemischen Arten sowie großer Landschaftsästhetik besitzt São Tomé einen hohen ökologischen Wert. Die Gesamtfläche des Primär- bzw. Altwaldes und des Sekundärwaldes beläuft sich auf ca. 50 %. Hinsichtlich einer schnell ansteigenden Einwohnerzahl auf São Tomé erhöht sich kontinuierlich der Landnutzungsdruck in Form von Walddegradation und Biodiversitätsgefährdung. Die methodischen Grundlagen der Forschungsarbeit basieren auf einem integrierten GIS- (Analyse bzw. Modellierung der LV) und Fernerkundungs-Konzept (LULC-Klassifikation). Das LV-Modell, gekennzeichnet durch eine linear-hierarchische Struktur, stützt sich auf bodenkundliche, topographische, fernerkundungsbasierte, statistische und infrastrukturelle Ausgangsdaten. Die Bewertungsanalyse erfolgt multifaktoriell mit einer anschließenden räumlichen Überlagerungsanalyse und gewichteter Summe. Die Ergebnisse sind nach der Intensitätsklassifizierung der LV räumlich-differenziert und geben Auskunft über die Intensität der Vulnerabilität in verschiedenen Landschaftsbereichen. Dadurch können Landschaftsabschnitte identifiziert werden, die für potentielle anthropogen verursachte Gefährdungen anfällig sind. Die gewonnene Information kann das Landmanagement optimieren und zum Biodiversitätsschutz auf São Tomé beitragen. Dank des exemplarischen Ansatzes ist dieses Konzept auch auf andere regional und klimatisch ähnliche tropische Systeme übertragbar. Darüber hinaus können die aus dem Modellansatz gewonnenen Erkenntnisse für die Bewertung der Vulnerabilität tropischer Ökosysteme auch zur Disaster Risk Reduction (DRR) beitragen.
Article
Full-text available
The recently available Landsat Analysis Ready Data (ARD) are provided as top of atmosphere (TOA) and atmospherically corrected (surface) reflectance tiled products and are designed to make the U.S. Landsat archive for the United States straightforward to use. In this study, the utility of ARD for 30 m percent tree cover mapping is demonstrated and the impact of different ARD processing levels on mapping accuracy examined. Five years of Landsat 5 and 7 ARD over 12 tiles encompassing Washington State are considered using an established bagged regression tree methodology and training data derived from Goddard LiDAR Hyperspectral & Thermal Imager (G-LiHT) data. Sensitivity to the amount of training data is examined with increasing mapping accuracy observed as more training data are used. Four processing levels of ARD are considered independently and the mapped results are compared: (i) TOA ARD; (ii) surface ARD; (iii) bidirectional reflectance distribution function (BRDF) adjusted atmospherically corrected ARD; and (iv) weekly composited BRDF adjusted atmospherically corrected ARD. The atmospherically corrected ARD provide marginally the highest mapping accuracies, although accuracy differences are negligible among the four (≤0.07% RMSE) when modest amounts of training data are used. The TOA ARD provide the most accurate maps compared to the other input data when only small amounts of training data are used, and the least accurate maps otherwise. The results are illustrated and the implications discussed.
Article
Full-text available
Google Earth Engine is a cloud-based platform for planetary-scale geospatial analysis that brings Google's massive computational capabilities to bear on a variety of high-impact societal issues including deforestation, drought, disaster, disease, food security, water management, climate monitoring and environmental protection. It is unique in the field as an integrated platform designed to empower not only traditional remote sensing scientists, but also a much wider audience that lacks the technical capacity needed to utilize traditional supercomputers or large-scale commodity cloud computing resources.
Article
Full-text available
The paper “Are Brazil deforesters avoiding detection?” recently published in Conservation Letters by Richards et al. 2016 has critical shortcomings and conclusions based on biased and not very robust analyses. Here we provide clarifications to some of the most critical points regarding the monitoring of land use changes in the Brazilian Amazon and related greenhouse emissions. Such clarifications are relevant to the readers of Conservation Letters and to a broader audience that rely on sound and robust science for a better management of environmental issues. This article is protected by copyright. All rights reserved
Article
Full-text available
The article “Are Brazil's Deforesters Avoiding Detection?” published recently in Conservation Letters offers a detailed analysis of the limits of PRODES, Brazil's best known system for monitoring deforestation in the Amazon rainforest. While the article provides a useful comparison of PRODES and two other monitoring systems, we strongly disagree with the authors’ suggestion that Brazil's monitoring systems are “antiquated and incomplete” and that they do not provide the basis for “transparently achieving Brazil‘s GHG [greenhouse gases] mitigation commitments” (Richards et al., 2016: 11). Richard's et al. ignore the existence of other monitoring systems developed by Brazilian government over the last decade. Thus, while PRODES still have a central role, the government has also at its disposal DETER to detect in near real-time large plots of forest degradation, DETEX for selective logging, DEGRAD for forest degradation, DETER-B for small plots of forest degradation and TerraClass for the monitoring of the increase and loss of secondary forests and other land uses (see Table 1 for a full list) (Almeida et al., 2016, Diniz et al., 2015). This article is protected by copyright. All rights reserved
Article
Full-text available
In tropical areas, pioneer occupation fronts steer the rapid expansion of deforestation, contributing to carbon emissions. Up-to-date carbon emission estimates covering the long-term development of such frontiers depend on the availability of high spatial–temporal resolution data. In this paper, we provide a detailed assessment of carbon losses from deforestation and potential forest degradation from fragmentation for one expanding frontier in the Brazilian Amazon. We focused on one of the Amazonia’s hot-spots of forest loss, the BR-163 highway that connects the high productivity agricultural landscapes in Mato Grosso with the exporting harbors of the Amazon. We used multi-decadal (1984–2012) Landsat-based time series on forested and non-forested area in combination with a carbon book-keeping model. We show a 36% reduction in 1984s biomass carbon stocks, which led to the emission of 611.5 TgCO2 between 1985 and 1998 (43.6 TgCO2 year⁻¹) and 959.8 TgCO2 over 1999–2012 (68.5 TgCO2 year⁻¹). Overall, fragmentation-related carbon losses represented 1.88% of total emissions by 2012, with an increasing relevance since 2004. We compared the Brazilian Space Agency deforestation assessment (PRODES) with our data and found that small deforestation polygons not captured by PRODES had increasing importance on estimated deforestation carbon losses since 2000. The comparative analysis improved the understanding of data-source-related uncertainties on carbon estimates and indicated disagreement areas between datasets that could be subject of future research. Furthermore, spatially explicit, annual deforestation and emission estimates like the ones derived from this study are important for setting regional baselines for REDD+ or similar payment for ecosystem services frameworks.
Article
Full-text available
Concerted political attention has focused on reducing deforestation, and this remains the cornerstone of most biodiversity conservation strategies. However, maintaining forest cover may not reduce anthropogenic forest disturbances, which are rarely considered in conservation programmes. These disturbances occur both within forests, including selective logging and wildfires, and at the landscape level, through edge, area and isolation effects. Until now, the combined effect of anthropogenic disturbance on the conservation value of remnant primary forests has remained unknown, making it impossible to assess the relative importance of forest disturbance and forest loss. Here we address these knowledge gaps using a large data set of plants, birds and dung beetles (1,538, 460 and 156 species, respectively) sampled in 36 catchments in the Brazilian state of Pará. Catchments retaining more than 69-80% forest cover lost more conservation value from disturbance than from forest loss. For example, a 20% loss of primary forest, the maximum level of deforestation allowed on Amazonian properties under Brazil's Forest Code, resulted in a 39-54% loss of conservation value: 96-171% more than expected without considering disturbance effects. We extrapolated the disturbance-mediated loss of conservation value throughout Pará, which covers 25% of the Brazilian Amazon. Although disturbed forests retained considerable conservation value compared with deforested areas, the toll of disturbance outside Pará's strictly protected areas is equivalent to the loss of 92,000-139,000 km(2) of primary forest. Even this lowest estimate is greater than the area deforested across the entire Brazilian Amazon between 2006 and 2015 (ref. 10). Species distribution models showed that both landscape and within-forest disturbances contributed to biodiversity loss, with the greatest negative effects on species of high conservation and functional value. These results demonstrate an urgent need for policy interventions that go beyond the maintenance of forest cover to safeguard the hyper-diversity of tropical forest ecosystems.
Article
Full-text available
Understanding spatial patterns of land use and land cover is essential for studies addressing biodiversity, climate change and environmental modeling as well as for the design and monitoring of land use policies. The aim of this study was to create a detailed map of land use land cover of the deforested areas of the Brazilian Legal Amazon up to 2008. Deforestation data from and uses were mapped with Landsat-5/TM images analysed with techniques, such as linear spectral mixture model, threshold slicing and visual interpretation, aided by temporal information extracted from NDVI MODIS time series. The result is a high spatial resolution of land use and land cover map of the entire Brazilian Legal Amazon for the year 2008 and corresponding calculation of area occupied by different land use classes. The results showed that the four classes of Pasture covered 62% of the deforested areas of the Brazilian Legal Amazon, followed by Secondary Vegetation with 21%. The area occupied by Annual Agriculture covered less than 5% of deforested areas; the remaining areas were distributed among six other land use classes. The maps generated from this project-called TerraClass-are available at INPE’s web site (http://www. inpe.br/cra/projetos_pesquisas/terraclass2008.php). © 2016, Instituto Nacional de Pesquisas da Amazonia. All rights reserved.
Article
Full-text available
Free and open access to the Landsat archive has enabled the implementation of national and global terrestrial monitoring projects. Herein, we summarize a project characterizing the change history of Canada’s forested ecosystems with a time series of data representing 1984–2012. Using the Composite2Change approach, we applied spectral trend analysis to annual best-available-pixel (BAP) surface reflectance image composites produced from Landsat TM and ETM+ imagery. A total of 73,544 images were used to produce 29 annual image composites, generating ∼400 TB of interim data products and resulting in ∼25 TB of annual gap-free reflectance composites and change products. On average, 10% of pixels in the annual BAP composites were missing data, with 86% of pixels having data gaps in two consecutive years or fewer. Change detection overall accuracy was 89%. Change attribution overall accuracy was 92%, with higher accuracy for stand-replacing wildfire and harvest. Changes were assigned to the correct year with an accuracy of 89%. Outcomes of this project provide baseline information and nationally consistent data source to quantify and characterize changes in forested ecosystems. The methods applied and lessons learned build confidence in the products generated and empower others to develop or refine similar satellite-based monitoring projects.
Article
Full-text available
Free and open access to satellite imagery and value-added data products have revolutionized the role of remote sensing in Earth system science. Nonetheless, rapid changes in the global environment pose challenges to the science community that are increasingly difficult to address using data from single satellite sensors or platforms due to the underlying limitations of data availability and tradeoffs that govern the design and implementation of currently existing sensors. Virtual constellations of planned and existing satellite sensors may help to overcome this limitation by combining existing observations to mitigate limitations of any one particular sensor. While multi-sensor applications are not new, the integration and harmonization of multi-sensor data is still challenging, requiring tremendous efforts of science and operational user communities. Defined by the Committee on Earth Observation Satellites (CEOS) as a “set of space and ground segment capabilities that operate in a coordinated manner to meet a combined and common set of Earth Observation requirements”, virtual constellations can principally be used to combine sensors with similar spatial, spectral, temporal, and radiometric characteristics. We extend this definition to also include sensors that are principally incompatible, because they are fundamentally different (for instance active versus passive remote sensing systems), but their combination is necessary and beneficial to achieve a specific monitoring goal. In this case, constellations are more likely to build upon the complementarity of resultant information products from these incompatible sensors rather than the raw physical measurements. In this communication, we explore the potential and possible limitations to be overcome regarding virtual constellations for terrestrial science applications, discuss potentials and limitations of various candidate sensors, and provide context on integration of sensors. Thematically, we focus on land-cover and land-use change (LCLUC), with emphasis given to medium spatial resolution (i.e., pixels sided 10 to 100 m) sensors, specifically as a complement to those onboard the Landsat series of satellites. We conclude that virtual constellations have the potential to notably improve observation capacity and thereby Earth science and monitoring programs in general. Various national and international parties have made notable and valuable progress related to virtual constellations. There is, however, inertia inherent to Earth observation programs, largely related to their complexity, as well as national interests, observation aims, and high system costs. Herein we define and describe virtual constellations, offer the science and applications information needs to offer context, provide the scientific support for a range of virtual constellation levels based upon applications readiness, capped by a discussion of issues and opportunities toward facilitating implementation of virtual constellations (in their various forms).
Article
Full-text available
The great success of the Brazilian deforestation programme " PRODES digital " has shown the importance of annual deforestation information for understanding and mitigating deforestation and its consequences in Brazil. However, there is a lack of similar information on deforestation for the 1990s and 1980s. Such maps are essential to understand deforestation frontier development and related carbon emissions. This study aims at extending the deforestation mapping record backwards into the 1990s and 1980s for one of the major deforestation frontiers in the Amazon. We use an image compositing approach to transform 2224 Landsat images in a spatially continuous and cloud free annual time series of Tasseled Cap Wetness metrics from 1984 to 2012. We then employ a random forest classifier to derive annual deforestation patterns. Our final deforestation map has an overall accuracy of 85% with half of the overall deforestation being detected before the year 2000. The results show for the first time detailed patterns of the expanding deforestation frontier before the 2000s. The high degree of automatization exhibits the great potential for mapping the whole Amazon biome using long-term and freely accessible remote sensing collections, such as the Landsat archive and forthcoming Sentinel-2 data.
Article
Full-text available
The Brazilian Legal Amazon (BLA), the largest global rainforest on earth, contains nearly 30% of the rainforest on earth. Given the regional complexity and dynamics, there are large government investments focused on controlling and preventing deforestation. The National Institute for Space Research (INPE) is currently developing five complementary BLA monitoring systems, among which the near real-time deforestation detection system (DETER) excels. DETER employs MODIS 250 m imagery and almost daily revisit, enabling an early warning system to support surveillance and control of deforestation. The aim of this paper is to present the methodology and results of the DETER based on AWIFS data, called DETER-B. Supported by 56 m images, the new system is effective in detecting deforestation smaller than 25 ha, concentrating 80% of its total detections and 45% of the total mapped area in this range. It also presents higher detection capability in identifying areas between 25 and 100 ha. The area estimation per municipality is statistically equal to those of the official deforestation data (PRODES) and allows the identification of degradation and logging patterns not observed with the traditional DETER system.
Article
Full-text available
Tower, ground-based and satellite observations indicate that tropical deforestation results in warmer, drier conditions at the local scale. Understanding the regional or global impacts of deforestation on climate, and ultimately on agriculture, requires modelling. General circulation models show that completely deforesting the tropics could result in global warming equivalent to that caused by burning of fossil fuels since 1850, with more warming and considerable drying in the tropics. More realistic scenarios of deforestation yield less warming and less drying, suggesting critical thresholds beyond which rainfall is substantially reduced. In regional, mesoscale models that capture topography and vegetation-based discontinuities, small clearings can actually enhance rainfall. At this smaller scale as well, a critical deforestation threshold exists, beyond which rainfall declines. Future agricultural productivity in the tropics is at risk from a deforestation-induced increase in mean temperature and the associated heat extremes and from a decline in mean rainfall or rainfall frequency. Through teleconnections, negative impacts on agriculture could extend well beyond the tropics.
Article
Full-text available
Atmospheric carbon dioxide records indicate that the land surface has acted as a strong global carbon sink over recent decades1, 2, with a substantial fraction of this sink probably located in the tropics3, particularly in the Amazon4. Nevertheless, it is unclear how the terrestrial carbon sink will evolve as climate and atmospheric composition continue to change. Here we analyse the historical evolution of the biomass dynamics of the Amazon rainforest over three decades using a distributed network of 321 plots. While this analysis confirms that Amazon forests have acted as a long-term net biomass sink, we find a long-term decreasing trend of carbon accumulation. Rates of net increase in above-ground biomass declined by one-third during the past decade compared to the 1990s. This is a consequence of growth rate increases levelling off recently, while biomass mortality persistently increased throughout, leading to a shortening of carbon residence times. Potential drivers for the mortality increase include greater climate variability, and feedbacks of faster growth on mortality, resulting in shortened tree longevity5. The observed decline of the Amazon sink diverges markedly from the recent increase in terrestrial carbon uptake at the global scale1, 2, and is contrary to expectations based on models6.
Article
Full-text available
We show that the vegetation canopy of the Amazon rainforest is highly sensitive to changes in precipitation patterns and that reduction in rainfall since 2000 has diminished vegetation greenness across large parts of Amazonia. Large-scale directional declines in vegetation greenness may indicate decreases in carbon uptake and substantial changes in the energy balance of the Amazon. We use improved estimates of surface reflectance from satellite data to show a close link between reductions in annual precipitation, El Niño southern oscillation events, and photosynthetic activity across tropical and subtropical Amazonia. We report that, since the year 2000, precipitation has declined across 69% of the tropical evergreen forest (5.4 million km2) and across 80% of the subtropical grasslands (3.3 million km2). These reductions, which coincided with a decline in terrestrial water storage, account for about 55% of a satellite-observed widespread decline in the normalized difference vegetation index (NDVI). During El Niño events, NDVI was reduced about 16.6% across an area of up to 1.6 million km2 compared with average conditions. Several global circulation models suggest that a rise in equatorial sea surface temperature and related displacement of the intertropical convergence zone could lead to considerable drying of tropical forests in the 21st century. Our results provide evidence that persistent drying could degrade Amazonian forest canopies, which would have cascading effects on global carbon and climate dynamics.
Article
Full-text available
When characterizing the processes that shape ecosystems, ecologists increasingly use the unique perspective offered by repeat observations of remotely sensed imagery. However, the concept of change embodied in much of the traditional remote-sensing literature was primarily limited to capturing large or extreme changes occurring in natural systems, omitting many more subtle processes of interest to ecologists. Recent technical advances have led to a fundamental shift toward an ecological view of change. Although this conceptual shift began with coarser-scale global imagery, it has now reached users of Landsat imagery, since these datasets have temporal and spatial characteristics appropriate to many ecological questions. We argue that this ecologically relevant perspective of change allows the novel characterization of important dynamic processes, including disturbances, long-term trends, cyclical functions, and feedbacks, and that these improvements are already facilitating our understanding of critical driving forces, such as climate change, ecological interactions, and economic pressures.
Article
Full-text available
The extent of the Amazon rainforest is projected to drastically decrease in future decades because of land-use changes. Previous climate modelling studies have found that the biogeophysical effects of future Amazonian deforestation will likely increase surface temperatures and reduce precipitation locally. However, the magnitude of these changes and the potential existence of tipping points in the underlying relationships is still highly uncertain. Using a regional climate model at a resolution of about 50 km over the South American continent, we perform four ERA-interim-driven simulations with prescribed land cover maps corresponding to present-day vegetation, two deforestation scenarios for the twenty-first century, and a totally-deforested Amazon case. In response to projected land cover changes for 2100, we find an annual mean surface temperature increase of 0.5°C over the Amazonian region and an annual mean decrease in rainfall of 0.17 mm/day compared to present-day conditions. These estimates reach 0.8°C and 0.22 mm/day in the total-deforestation case. We also compare our results to those from 28 previous (regional and global) climate modelling experiments. We show that the historical development of climate models did not modify the median estimate of the Amazonian climate sensitivity to deforestation, but led to a reduction of its uncertainty. Our results suggest that the biogeophysical effects of deforestation alone are unlikely to lead to a tipping point in the evolution of the regional climate under present-day climate conditions. However, the conducted synthesis of the literature reveals that this behaviour may be model-dependent, and the greenhouse gas-induced climate forcing and biogeochemical feedbacks should also be taken into account to fully assess the future climate of this region.
Article
Full-text available
Information on the changing land surface is required at high spatial resolutions as many processes cannot be resolved using coarse resolution data. Deriving such information over large areas for Landsat data, however, still faces numerous challenges. Image compositing offers great potential to circumvent such shortcomings. We here present a compositing algorithm that facilitates creating cloud free, seasonally and radiometrically consistent datasets from the Landsat archive. A parametric weighting scheme allows for flexibly utilizing different pixel characteristics for optimized compositing. We describe in detail the development of three parameter decision functions: acquisition year, day of year and distance to clouds. Our test site covers 42 Landsat footprints in Eastern Europe and we produced three annual composites. We evaluated seasonal and annual consistency and compared our composites to BRDF normalized MODIS reflectance products. Finally, we also evaluated how well the composites work for land cover mapping. Results prove that our algorithm allows for creating seasonally consistent large area composites. Radiometric correspondence to MODIS was high (up to R2 > 0.8), but varied with land cover configuration and selected image acquisition dates. Land cover mapping yielded promising results (overall accuracy 72%). Class delineations were regionally consistent with minimal effort for training data. Class specific accuracies increased considerably (~10%) when spectral metrics were incorporated. Our study highlights the value of compositing in general and for Landsat data in particular, allowing for regional to global LULCC mapping at high spatial resolutions.
Article
Full-text available
Forests in Flux Forests worldwide are in a state of flux, with accelerating losses in some regions and gains in others. Hansen et al. (p. 850 ) examined global Landsat data at a 30-meter spatial resolution to characterize forest extent, loss, and gain from 2000 to 2012. Globally, 2.3 million square kilometers of forest were lost during the 12-year study period and 0.8 million square kilometers of new forest were gained. The tropics exhibited both the greatest losses and the greatest gains (through regrowth and plantation), with losses outstripping gains.
Article
Full-text available
A new method called Fmask (Function of mask) for cloud and cloud shadow detection in Landsat imagery is provided. Landsat Top of Atmosphere (TOA) reflectance and Brightness Temperature (BT) are used as inputs. Fmask first uses rules based on cloud physical properties to separate Potential Cloud Pixels (PCPs) and clear-sky pixels. Next, a normalized temperature probability, spectral variability probability, and brightness probability are combined to produce a probability mask for clouds over land and water separately. Then, the PCPs and the cloud probability mask are used together to derive the potential cloud layer. The darkening effect of the cloud shadows in the Near Infrared (NIR) Band is used to generate a potential shadow layer by applying the flood-fill transformation. Subsequently, 3D cloud objects are determined via segmentation of the potential cloud layer and assumption of a constant temperature lapse rate within each cloud object. The view angle of the satellite sensor and the illuminating angle are used to predict possible cloud shadow locations and select the one that has the maximum similarity with the potential cloud shadow mask. If the scene has snow, a snow mask is also produced. For a globally distributed set of reference data, the average Fmask overall cloud accuracy is as high as 96.4%. The goal is development of a cloud and cloud shadow detection algorithm suitable for routine usage with Landsat images.
Article
Full-text available
Techniques based on multi-temporal, multi-spectral, satellite-sensor- acquired data have demonstrated potential as a means to detect, identify, map and monitor ecosystem changes, irrespective of their causal agents. This review paper, which summarizes the methods and the results of digital change detection in the optical/infrared domain, has as its primary objective a synthesis of the state of the art today. It approaches,digital change,detection from,three angles. First, the different perspectives from which the variability in ecosystems and the change,events have been dealt with are summarized.,Change,detection between pairs of images,(bi-temporal) as well as between,time profiles of imagery,derived indicators (temporal trajectories), and, where relevant, the appropriate choices for digital imagery acquisition timing and change interval length definition, are discussed. Second, pre-processing routines either to establish a more direct linkage between remote sensing data and biophysical phenomena, or to temporally mosaic imagery and extract time profiles, are reviewed. Third, the actual change,detection,methods,themselves,are categorized,in an analytical framework and critically evaluated. Ultimately, the paper highlights how some of these methodological,aspects are being,fine-tuned as this review,is being written, and we summarize the new developments that can be expected in the near future. The review,highlights the high complementarity,between,different change,detection methods.
Article
Full-text available
Climate change and rapidly escalating global demand for food, fuel, fibre and feed present seemingly contradictory challenges to humanity. Can greenhouse gas (GHG) emissions from land-use, more than one-fourth of the global total, decline as growth in land-based production accelerates? This review examines the status of two major international initiatives that are designed to address different aspects of this challenge. REDD+ is an emerging policy framework for providing incentives to tropical nations and states that reduce their GHG emissions from deforestation and forest degradation. Market transformation, best represented by agricultural commodity roundtables, seeks to exclude unsustainable farmers from commodity markets through international social and environmental standards for farmers and processors. These global initiatives could potentially become synergistically integrated through (i) a shared approach for measuring and favouring high environmental and social performance of land use across entire jurisdictions and (ii) stronger links with the domestic policies, finance and laws in the jurisdictions where agricultural expansion is moving into forests. To achieve scale, the principles of REDD+ and sustainable farming systems must be embedded in domestic low-emission rural development models capable of garnering support across multiple constituencies. We illustrate this potential with the case of Mato Grosso State in the Brazilian Amazon.
Article
Full-text available
Timely and accurate change detection of Earth’s surface features is extremely important for understanding relationships and interactions between human and natural phenomena in order to promote better decision making. Remote sensing data are primary sources extensively used for change detection in recent decades. Many change detection techniques have been developed. This paper summarizes and reviews these techniques. Previous literature has shown that image differencing, principal component analysis and post-classification comparison are the most common methods used for change detection. In recent years, spectral mixture analysis, artificial neural networks and integration of geographical information system and remote sensing data have become important techniques for change detection applications. Different change detection algorithms have their own merits and no single approach is optimal and applicable to all cases. In practice, different algorithms are often compared to find the best change detection results for a specific application. Research of change detection techniques is still an active topic and new techniques are needed to effectively use the increasingly diverse and complex remotely sensed data available or projected to be soon available from satellite and airborne sensors. This paper is a comprehensive exploration of all the major change detection approaches implemented as found in the literature.