Forest Mapping Research Articles

Characterizing historical transformation trajectories of the forest landscape in Rome's metropolitan area (Italy) for effective planning of sustainability goals

AbstractWith the aim at developing a landscape dynamics framework for environmental planning and management and testing the effectiveness of protected areas in achieving the 2030 Agenda of the United Nations sustainability goals, we characterized the historical transformation trajectories of forest area changes from 1936 to 2010 in the Metropolitan City of Rome Capital (Italy). Remote sensing‐based products coupled with landscape pattern metrics and fragmentation analysis have been implemented, comparing different historical forest maps. The results show a remarkable forest area gain – from 17.6% to 25.5% – thanks to 68,299 ha of recently established forest. Statistical descriptors showed that the highest relative gain occurred in mountain zones, confirming a wide European forest recovery pattern in marginal areas from past deforestation and overexploitation. Deforestation mainly occurred in the flat and hilly areas where almost 26,000 ha of forests were lost since 1936. In summary, two main forest landscape dynamics were reconstructed: (I) the increase of forest cover fragmentation in the lowland areas; and (II) the rise in the forest area in the interior sectors of the mountain landscape, mainly within protected areas. Restoring the forest ecosystem's bioecological integrity has been highlighted as an urgent action for biodiversity conservation and carbon mitigation. In lowland areas, the study revealed the urgent need to establish new protected areas and rewilding spaces as landscape metrics are relatively below the sustainability targets for healthy forest ecosystems. The proposed framework can be used for testing the effectiveness of environmental planning and management in other forest landscapes to achieve the Agenda 2030 goals.

Informing forest conservation planning with detailed human footprint data for Argentina

Conserving the remaining wildest forests is a top priority for conservation, and human footprint maps are a practical way to identify wild areas. However, available global assessments of wild areas are too coarse for land use decisions, especially in countries with high deforestation rates, such as Argentina. Our main goal was to map the human footprint in Argentina’s forested areas to improve conservation planning at regional and country levels. Specifically, we quantified the level of human influence on the environment and mapped the wildest native forests (i) across forest regions, and (ii) in the different land-use categories of the National Forest Plan, which is a key policy instrument for conserving the nation’s native forests through zoning, and (iii) identified wildest forests that are at risk due to human activities. We analyzed detailed spatial data on settlements, transportation, energy, and land use change, and estimated the areal extent to which these various human activities disrupt natural processes. We defined pixels with human footprint index of zero as wildest areas. We found that a substantial portion (43%) of Argentina’s forested area remains wild, which suggests there are opportunities for conservation. However, levels of human influence varied substantially among forest regions, and Atlantic and Chaco forests have the highest levels of human influence. Further, we found that the National Forest Plan does not conserve the wildest forests of the nation, as most (78%) of the wildest native forests are located in zones that allow silvopasture, timber production, and/or forest conversion to crops, thus potentially threatening biodiversity in these areas. Our map of wildest forests is an important, but first, step in identifying wildland forests in Argentina, as available spatial data layers of human activities capture many, but not all, human influences on forests. For instance, small human features, like certain rural roads, trails, and rural settlements exist in our wildest areas. Our study provides new datasets to assist land use planners and conservationists, and identifies areas for conservation attention in Argentina. More broadly, our analyses highlight the value of detailed human footprint data to support conservation decisions in forest landscapes.

Object-based random forest wetland mapping in Conne River, Newfoundland, Canada

The Conne River watershed is dominated by wetlands that provide valuable ecosystem services, including contributing to the survivability and propagation of Atlantic salmon, an important subsistence species that has shown a dramatic decline over the past 30 years. To better understand and improve the management of the watershed, and in turn, the Atlantic salmon, a wetland inventory of the area is developed using advanced remote sensing methods including field-collected data, object-based image analysis of Sentinel-1, Sentinel-2, and digital elevation model Earth observation data. The resulting classification maps consisted of bog, fen, swamp, marsh, and open water wetlands with an overall accuracy of 92% and a kappa coefficient of 0.916. Among wetland classes, user and producer accuracies range between 84% and 100%. Results show the dominance of peatland wetlands such as bog and fen, and the relative rareness of marsh wetlands.

Mapping and Monitoring Zero-Deforestation Commitments.

A growing number of companies have announced zero-deforestation commitments (ZDCs) to eliminate commodities produced at the expense of forests from their supply chains. Translating these aspirational goals into forest conservation requires forest mapping and monitoring (M&M) systems that are technically adequate and therefore credible, salient so that they address the needs of decision makers, legitimate in that they are fair and unbiased, and scalable over space and time. We identify 12 attributes of M&M that contribute to these goals and assess how two prominent ZDC programs, the Amazon Soy Moratorium and the High Carbon Stock Approach, integrate these attributes into their M&M systems. These programs prioritize different attributes, highlighting fundamental trade-offs in M&M design. Rather than prescribe a one-size-fits-all solution, we provide policymakers and practitioners with guidance on the design of ZDC M&M systems that fit their specific use case and that may contribute to more effective implementation of ZDCs.

Improved forest fire spread mapping by developing custom fire fuel models in replanted forests in Hyrcanian forests, Iran

Aim of the study:Forest fuel classification and characterization is a critical factor in wildfire management. The main purpose of this study was to develop custom fuel models for accurately mapping wildfire spread compared to standard models.Area of study: The study was conducted at a replanted forest dominated by coniferous species, in the Arabdagh region,GolestanProvince, northernIran.Material and methods: Six custom fuel models were developed to characterize the main vegetation types in the study area. Fuel samples were collected from 49 randomly selected plots. In each plot, the fuel load of 1-hr, 10-hr, 100-hr, 1000-hr, live herbs, live woody plants, surface area volume ratio, and fuel depth were estimated using the Fuel Load (FL) sampling method along three transects. Canopy fuel load was calculated for each fuel model. The performance of the custom fuel models versus standard fuel models on wildfire behavior simulations was compared using the FlamMap MTT simulator.Main results: The results showed that, despite the similarity in the burned area between observed and modeled fires, the custom fuel models produced an increase in simulation accuracy. Compared to the observed fire, simulation results did not give realistic results to the crown fire. The simulation using standard fuel models did not result in crown fire, while the simulation using custom fuel models showed a moderate rate of crown fire with a Kappa coefficient of 0.54.Research highlights: The results demonstrated the importance of developing custom fuel models to simulate wildfire maps with higher accuracy for wildfire risk management.Keywords: custom fuel model; FlamMap; replantation; vegetation type; wildfire behavior.

Estimating average tree height in Xixiaoshan Forest Farm, Northeast China based on Sentinel-1 with Sentinel-2A data

Forest resource survey is important for the sustainable development of forest ecosystem in China. The average tree height is a main structural parameter of forest resource survey, and also one of the key parameters with greatest difficulty to obtain. The purpose of this study was to explore the potential of joint active and passive remote sensing technology in estimating forest average height. Taking Xixiaoshan Forest Farm in Linjiang City of Jilin Province as the research area, we used Sentinel-1 SAR and Sentinel-2A data, extracted two backscatter coefficients and eight texture information of Sentinel-1, ten spectral bands and texture information of Sentinel-2A and eleven vegetation index variables, constructed five groups of average tree height estimation models based on above variables and fusion of four variables by multiple linear regression method. We further evaluated the influence of each variable on the inversion accuracy. The results showed that the texture information extracted from the Sentinel-2A spectral band of a single data source variable had a better modeling effect and could be used as effective data to estimate the average tree height. The height estimation model of the integrated four variables was optimal, with a R2 vaule of 0.56, a root mean square error of leave-one-out cross-validation of 2.92 m, and a relative root mean square error of leave-one-out cross-validation of 21.5%. The forest average height model based on Sentinel-1 and Sentinel-2a characteristic variables could improve the estimation accuracy of forest height, which could be used for regional forest average height estimation and mapping.

Identification of CDK2-Related Immune Forecast Model and ceRNA in Lung Adenocarcinoma, a Pan-Cancer Analysis.

BackgroundTumor microenvironment (TME) plays important roles in different cancers. Our study aimed to identify molecules with significant prognostic values and construct a relevant Nomogram, immune model, competing endogenous RNA (ceRNA) in lung adenocarcinoma (LUAD).Methods“GEO2R,” “limma” R packages were used to identify all differentially expressed mRNAs from Gene Expression Omnibus (GEO) and The Cancer Genome Atlas (TCGA) databases. Genes with P-value <0.01, LogFC>2 or <-2 were included for further analyses. The function analysis of 250 overlapping mRNAs was shown by DAVID and Metascape software. By UALCAN, Oncomine and R packages, we explored the expression levels, survival analyses of CDK2 in 33 cancers. “Survival,” “survminer,” “rms” R packages were used to construct a Nomogram model of age, gender, stage, T, M, N. Univariate and multivariate Cox regression were used to establish prognosis-related immune forecast model in LUAD. CeRNA network was constructed by various online databases. The Genomics of Drug Sensitivity in Cancer (GDSC) database was used to explore correlations between CDK2 expression and IC50 of anti-tumor drugs.ResultsA total of 250 differentially expressed genes (DEGs) were identified to participate in many cancer-related pathways, such as activation of immune response, cell adhesion, migration, P13K-AKT signaling pathway. The target molecule CDK2 had prognostic value for the survival of patients in LUAD (P = 5.8e-15). Through Oncomine, TIMER, UALCAN, PrognoScan databases, the expression level of CDK2 in LUAD was higher than normal tissues. Pan-cancer analysis revealed that the expression, stage and survival of CDK2 in 33 cancers, which were statistically significant. Through TISIDB database, we selected 13 immunodepressants, 21 immunostimulants associated with CDK2 and explored 48 genes related to these 34 immunomodulators in cBioProtal database (P < 0.05). Gene Set Enrichment Analysis (GSEA) and Metascape indicated that 49 mRNAs were involved in PUJANA ATM PCC NETWORK (ES = 0.557, P = 0, FDR = 0), SIGNAL TRANSDUCTION (ES = –0.459, P = 0, FDR = 0), immune system process, cell proliferation. Forest map and Nomogram model showed the prognosis of patients with LUAD (Log-Rank = 1.399e-08, Concordance Index = 0.7). Cox regression showed that four mRNAs (SIT1, SNAI3, ASB2, and CDK2) were used to construct the forecast model to predict the prognosis of patients (P < 0.05). LUAD patients were divided into two different risk groups (low and high) had a statistical significance (P = 6.223e-04). By “survival ROC” R package, the total risk score of this prognostic model was AUC = 0.729 (SIT1 = 0.484, SNAI3 = 0.485, ASB2 = 0.267, CDK2 = 0.579). CytoHubba selected ceRNA mechanism medicated by potential biomarkers, 6 lncRNAs-7miRNAs-CDK2. The expression of CDK2 was associated with IC50 of 89 antitumor drugs, and we showed the top 20 drugs with P < 0.05.ConclusionIn conclusion, our study identified CDK2 related immune forecast model, Nomogram model, forest map, ceRNA network, IC50 of anti-tumor drugs, to predict the prognosis and guide targeted therapy for LUAD patients.

Classification of forest fire areas using machine learning algorithm

Forest fires that occur will cause various kinds of problems, both in terms of health, such as smoke that can interfere with the respiratory system, in terms of the economy such as the economic wheel cannot run as usual, in terms of the environment can damage the surrounding environment and the environment that is missed by smoke, and other disasters. Forest fires can also have an impact on the costs that will be incurred to resolve the problems that arise due to forest fires, so research is needed to find out and measure the area affected by forest fires that burned in the range of 1980 - 2019 using a dataset of approximately 10,000. The target in this research is to be able to generate the best percentage scenario and find out the model of using the algorithm used to explore the algorithm in the Machine Learning method for the model for estimating the area of forest fires, namely the Siak Kampar Peninsula in Riau Province. In this study, 7 parameters were used to create a forest and land fire hazard map, namely weather temperature, Burned Area Density, hotspot density, wind speed, land cover type, rainfall, and land use. The seven parameters will be searched for accuracy results using the Classification method with Machine Learning algorithms, including Naïve Bayes, SVM, and K-Nearest Neighbor (K-NN). In this study, comparisons were made to obtain the best algorithm for estimating forest fire areas. By generating each algorithm is 71.72% for the Naïve Bayes algorithm, 75.00% for the SVM algorithm, and 64.71% for the K-NN algorithm.

Multi-Model Approaches to the Spatialization of Tree Vitality Surveys: Constructing a National Tree Vitality Map

It is essential to maintain the health of forests so that they are protected against a diverse range of stressors and show improved resilience. An area-based forest health map is required for efficient forest management on a national scale however, most national forest inventories are based on in-situ observations. This study examined methodologies to establish an area-based map on tree vitality grade using field survey data, particularly that containing information on several trees at one point. The forest health monitoring dataset of the Republic of Korea was used in combination with 37 satellite-based environmental predictors. Four methods were considered: Multinomial logistic regression (MLR), random forest classification (RF), indicator kriging (IK), and multi-model ensemble (MME) approaches using species distribution models. The MLR and RF produced biased results, whereby almost all regions were classified as first grade; the spatialization results of these methods were considered inappropriate for forest management. The maps produced using the IK and MME methods improved the distinctions between the distributions of five grades compared to the previous two methodologies however, the MME method produced better results, reliably reflecting topographical and climatic characteristics. Comparisons with the vegetation condition index and bioclimate vulnerability index also emphasized the usefulness of the MME. This study is particularly relevant to the national forest managers who struggle to find the most effective forest monitoring and management strategies. Suggestions to improve spatialization of field survey data are further discussed.

Burned Area Mapping over the Southern Cape Forestry Region, South Africa Using Sentinel Data within GEE Cloud Platform

Planted forests in South Africa have been affected by an increasing number of economically damaging fires over the past four decades. They constitute a major threat to the forestry industry and account for over 80% of the country’s commercial timber losses. Forest fires are more frequent and severe during the drier drought conditions that are typical in South Africa. For proper forest management, accurate detection and mapping of burned areas are required, yet the exercise is difficult to perform in the field because of time and expense. Now that ready-to-use satellite data are freely accessible in the cloud-based Google Earth Engine (GEE), in this study, we exploit the Sentinel-2-derived differenced normalized burned ratio (dNBR) to characterize burn severity areas, and also track carbon monoxide (CO) plumes using Sentinel-5 following a wildfire that broke over the southeastern coast of the Western Cape province in late October 2018. The results showed that 37.4% of the area was severely burned, and much of it occurred in forested land in the studied area. This was followed by 24.7% of the area that was burned at a moderate-high level. About 15.9% had moderate-low burned severity, whereas 21.9% was slightly burned. Random forests classifier was adopted to separate burned class from unburned and achieved an overall accuracy of over 97%. The most important variables in the classification included texture, NBR, and the NIR bands. The CO signal sharply increased during fire outbreaks and marked the intensity of black carbon over the affected area. Our study contributes to the understanding of forest fire in the dynamics over the Southern Cape forestry landscape. Furthermore, it also demonstrates the usefulness of Sentinel-5 for monitoring CO. Taken together, the Sentinel satellites and GEE offer an effective tool for mapping fires, even in data-poor countries.

Estimating Primary Forest Attributes and Rare Community Characteristics Using Unmanned Aerial Systems (UAS): An Enrichment of Conventional Forest Inventories

The techniques for conducting forest inventories have been established over centuries of land management and conservation. In recent decades, however, compelling new tools and methodologies in remote sensing, computer vision, and data science have offered innovative pathways for enhancing the effectiveness and comprehension of these sampling designs. Now with the aid of Unmanned Aerial Systems (UAS) and advanced image processing techniques, we have never been closer to mapping forests at field-based inventory scales. Our research, conducted in New Hampshire on complex mixed-species forests, used natural color UAS imagery for estimating individual tree diameters (diameter at breast height (dbh)) as well as stand level estimates of Basal Area per Hectare (BA/ha), Quadratic Mean Diameter (QMD), Trees per Hectare (TPH), and a Stand Density Index (SDI) using digital photogrammetry. To strengthen our understanding of these forests, we also assessed the proficiency of the UAS to map the presence of large trees (i.e., &gt;40 cm in diameter). We assessed the proficiency of UAS digital photogrammetry for identifying large trees in two ways: (1) using the UAS estimated dbh and the 40 cm size threshold and (2) using a random forest supervised classification and a combination of spectral, textural, and geometric features. Our UAS-based estimates of tree diameter reported an average error of 19.7% to 33.7%. At the stand level, BA/ha and QMD were overestimated by 42.18% and 62.09%, respectively, while TPH and SDI were underestimated by 45.58% and 3.34%. When considering only stands larger than 9 ha however, the overestimation of BA/ha at the stand level dropped to 14.629%. The overall classification of large trees, using the random forest supervised classification achieved an overall accuracy of 85%. The efficiency and effectiveness of these methods offer local land managers the opportunity to better understand their forested ecosystems. Future research into individual tree crown detection and delineation, especially for co-dominant or suppressed trees, will further support these efforts.

Development and Application of Earth Observation Based Machine Learning Methods for Characterizing Forest and Land Cover Change in Dilijan National Park of Armenia between 1991 and 2019

Dilijan National Park is one of the most important national parks of Armenia, established in 2002 to protect its rich biodiversity of flora and fauna and to prevent illegal logging. The aim of this study is to provide first, a mapping of forest degradation and deforestation, and second, of land cover/land use changes every 5 years over a 28-year monitoring cycle from 1991 to 2019, using Sentinel-2 and Landsat time series and Machine Learning methods. Very High Spatial Resolution imagery was used for calibration and validation purposes of forest density modelling and related changes. Correlation coefficient R2 between forest density map and reference values ranges from 0.70 for the earliest epoch to 0.90 for the latest one. Land cover/land use classification yield good results with most classes showing high users’ and producers’ accuracies above 80%. Although forest degradation and deforestation which initiated about 30 years ago was restrained thanks to protection measures, anthropogenic pressure remains a threat with the increase in settlements, tourism, or agriculture. This case study can be used as a decision-support tool for the Armenian Government for sustainable forest management and policies and serve as a model for a future nationwide forest monitoring system.

Mapping Forest Height and Aboveground Biomass by Integrating ICESat‐2, Sentinel‐1 and Sentinel‐2 Data Using Random Forest Algorithm in Northwest Himalayan Foothills of India

AbstractThe present study aims to map forest canopy height by integrating ICESat‐2 and Sentinel‐1 data and investigate the effect of integrating forest canopy height information with Sentinel‐2 data‐derived spectral variables on the prediction of spatial distribution of forest aboveground biomass (AGB). Random forest (RF) algorithm was used to develop forest canopy height and AGB models. It was observed that ICESat‐2 and Sentinel‐1 based model was able to predict forest canopy height with R2 = 0.84 and %RMSE = 4.48%. Two forest AGB models were developed, with only spectral variables and by incorporating forest height information with spectral variables. The results reflected that incorporation of forest canopy height in the forest AGB model improved the accuracy of the AGB predictions (R2 = 0.83, %RMSE = 4.64%). The study presents a comprehensive methodology for mapping forest canopy height and AGB.

GIS in the jungle: Experiential Environmental Education (EEE) in Panama.

A University of Redlands travel course to Panama builds on Monty Hempel’s legacy in environmental studies, Experiential Environmental Education, and geographic information systems (GIS) instruction at the University of Redlands. It follows his experience developing social relationships in Panama, and nearly 20 years leading an EEE travel course to Palau. This Panama course, nicknamed “GIS in the Jungle,” tasks students with activities such as trail mapping, gathering original data with unmanned aerial vehicles (UAV), landcover mapping of pastures and forests, and visualizing the distribution of above-ground forest carbon content. These GIS activities help students construct an understanding of rainforest conservation that integrates their experiences living, hiking, and collecting original data in that environment. UAV and GIS also provided a platform for interacting with indigenous peoples struggling to defend their rainforest territory from colonist deforestation. As Monty Hempel observed about this course, students use GIS and UAV to integrate their direct experience with ground-level fieldwork in the rainforest with a birds-eye view of the interlocking ecosystems and human impacts. Combined with direct rainforest experiences, GIS projects help students develop an understanding of nature’s interlocking systems and the interdependence of life on Earth.

A Novel Spatial Simulation Method for Mapping the Urban Forest Carbon Density in Southern China by the Google Earth Engine

Urban forest is an important component of terrestrial ecosystems and is highly related to global climate change. However, because of complex city landscapes, deriving the spatial distribution of urban forest carbon density and conducting accuracy assessments are difficult. This study proposes a novel spatial simulation method, optimized geographically weighted logarithm regression (OGWLR), using Landsat 8 data acquired by the Google Earth Engine (GEE) and field survey data to map the forest carbon density of Shenzhen city in southern China. To verify the effectiveness of the novel method, multiple linear regression (MLR), k-nearest neighbors (kNN), random forest (RF) and geographically weighted regression (GWR) models were established for comparison. The results showed that OGWLR achieved the highest coefficient of determination (R2 = 0.54) and the lowest root mean square error (RMSE = 13.28 Mg/ha) among all estimation models. In addition, OGWLR achieved a more consistent spatial distribution of carbon density with the actual situation. The carbon density of the forests in the study area was large in the central and western regions and coastal areas and small in the building and road areas. Therefore, this method can provide a new reference for urban forest carbon density estimation and mapping.

Forest Aboveground Biomass Estimation and Mapping through High-Resolution Optical Satellite Imagery—A Literature Review

This paper provides a comprehensive literature review on forest aboveground biomass (AGB) estimation and mapping through high-resolution optical satellite imagery (≤5 m spatial resolution). Based on the literature review, 44 peer-reviewed journal articles were published in 15 years (2004–2019). Twenty-one studies were conducted in Asia, eight in North America and Africa, five in South America, and four in Europe. This review article gives a glance at the published methodologies for AGB prediction modeling and validation. The literature review suggested that, along with the integration of other sensors, QuickBird, WorldView-2, and IKONOS satellite images were most widely used for AGB estimations, with higher estimation accuracies. All studies were grouped into six satellite-derived independent variables, including tree crown, image textures, tree shadow fraction, canopy height, vegetation indices, and multiple variables. Using these satellite-derived independent variables, most of the studies used linear regression (41%), while 30% used linear multiple regression and 18% used non-linear (machine learning) regression, while very few (11%) studies used non-linear (multiple and exponential) regression for estimating AGB. In the context of global forest AGB estimations and monitoring, the advantages, strengths, and limitations were discussed to achieve better accuracy and transparency towards the performance-based payment mechanism of the REDD+ program. Apart from technical limitations, we realized that very few studies talked about real-time monitoring of AGB or quantifying AGB change, a dimension that needs exploration.

Spatially explicit changes in forest biomass carbon of China over the past 4 decades: Coupling long-term inventory and remote sensing data

Quantification of the spatial pattern of forest carbon (C) sinks in high resolution is helpful to reveal the factors that affect the C cycle and provides valuable information for developing sustainable forest management policies. Here we developed a method using the data of long-term forest inventories (1977–2018) and spatially-explicit remotely sensed information from land-use maps and the Normalized Difference Vegetation Index (NDVI) datasets, to estimate the spatial and temporal variation of forest biomass C in China. At first, we calculated forest biomass C stocks using the refined Continuous Biomass Expansion Factor (CBEF) model with parameters for each forest type based on eight national forest inventories. Secondly, based on multi-temporal land-use remote sensing and national forest inventory datasets, we obtained forest coverage datasets with high resolution (1 km*1 km). Thirdly, we downscaled the forest biomass C density using the calibrated forest coverage maps and the maximum NDVI values derived from GIMMS-NDVI3g imagery. Our results showed that China's forest functioned as a C sink of 3777.73 Tg C, and the C density of forest stands increased from 35.41 Mg C ha−1 during 1977–1981 to 43.95 Mg C ha−1 during 2014–2018. In addition, the validation results for most of the provinces based on published inventory estimates during the eight periods showed that the forest area at the pixel scale was successfully calibrated. From this, we produced the maps with a finer resolution for a series of spatially continuous forest biomass carbon density distribution and carbon sinks. Notably eight major forest projects have accounted for 44%–51% of the forest C stocks added in China from 1977 to 2018. Our research provides new insights for understanding and monitoring the spatiotemporal variations in of forest biomass and key information to support the development of new afforestation policies moving forward.

Lessons Learned While Implementing a Time-Series Approach to Forest Canopy Disturbance Detection in Nepal

While deforestation has traditionally been the focus for forest canopy disturbance detection, forest degradation must not be overlooked. Both deforestation and forest degradation influence carbon loss and greenhouse gas emissions and thus must be included in activity data reporting estimates, such as for the Reduced Emissions from Deforestation and Degradation (REDD+) program. Here, we report on efforts to develop forest degradation mapping capacity in Nepal based on a pilot project in the country’s Terai region, an ecologically complex physiographic area. To strengthen Nepal’s estimates of deforestation and forest degradation, we applied the Continuous Degradation Detection (CODED) algorithm, which uses a time series of the Normalized Degradation Fraction Index (NDFI) to monitor forest canopy disturbances. CODED can detect low-grade degradation events and provides an easy-to-use graphical user interface in Google Earth Engine (GEE). Using an iterative process, we were able to create a model that provided acceptable accuracy and area estimates of forest degradation and deforestation in Terai that can be applied to the whole country. We found that between 2010 and 2020, the area affected by disturbance was substantially larger than the deforested area, over 105,650 hectares compared to 2753 hectares, respectively. Iterating across multiple parameters using the CODED algorithm in the Terai region has provided a wealth of insights not only for detecting forest degradation and deforestation in Nepal in support of activity data estimation but also for the process of using tools like CODED in applied settings. We found that model performance, measured using producer’s and user’s accuracy, varied dramatically based on the model parameters specified. We determined which parameters most altered the results through an iterative process; those parameters are described here in depth. Once CODED is combined with the description of each parameter and how it affects disturbance monitoring in a complex environment, this degradation-sensitive detection process has the potential to be highly attractive to other developing countries in the REDD+ program seeking to accurately monitor their forests.

Instance segmentation of fallen trees in aerial color infrared imagery using active multi-contour evolution with fully convolutional network-based intensity priors

Over the last several years, semantic image segmentation based on deep neural networks has been greatly advanced. On the other hand, single-instance segmentation still remains a challenging problem. In this paper, we introduce a framework for segmenting instances of a common object class by multiple active contour evolution over semantic segmentation maps of images obtained through fully convolutional networks. The contour evolution is cast as an energy minimization problem, where the aggregate energy functional incorporates a data fit term, an explicit shape model, and accounts for object overlap. Efficient solution neighborhood operators are proposed, enabling optimization through metaheuristics such as simulated annealing. We instantiate the proposed framework in the context of segmenting individual fallen stems from high-resolution aerial multispectral imagery, providing problem-specific energy potentials. We validated our approach on 3 real-world scenes of varying complexity, using 730 manually labeled polygon outlines as ground truth. The test plots were situated in regions of the Bavarian Forest National Park, Germany, which sustained a heavy bark beetle infestation. Evaluations were performed on both the polygon and line segment level, showing that the multi-contour segmentation can achieve up to 0.93 precision and 0.82 recall. An improvement of up to 7 percentage points (pp) in recall and 6 in precision compared to an iterative sample consensus line segment detection baseline was achieved. Despite the simplicity of the applied shape parametrization, an explicit shape model incorporated into the energy function improved the results by up to 4 pp of recall. Finally, we show the importance of using a high-quality semantic segmentation method (e.g. U-net) as the basis for individual stem detection, as the quality of the results degraded dramatically in our baseline experiment utilizing a simpler method. Our method is a step towards increased accessibility of automatic fallen tree mapping in forests, due to higher cost efficiency of aerial imagery acquisition compared to laser scanning. The precise fallen tree maps could be further used as a basis for plant and animal habitat modeling, studies on carbon sequestration as well as soil quality in forest ecosystems.

Diagnostic Accuracy of Procalcitonin for Bacterial Infection in Liver Failure: A Meta-Analysis.

The purpose of our studies was to systematically assess the accuracy and clinical value of plasma calcitonin in patients with liver failure complicated with bacterial infection. In this study, we included prospective observational studies or randomized controlled trials on PCT. The quality of the studies was evaluated using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool. Heterogeneity, pooled diagnostic odds ratio (DOR), pooled sensitivity, pooled specificity, pooled positive likelihood ratio, pooled negative likelihood ratio, the area under the summary receiver operating characteristic curve (SROC), and metaregression analysis were performed using Stata16.0 software. Consequently, the studies revealed substantial heterogeneity (I2 = 96, 95% confidence interval (95% CI) = 94–99). The results of meta-analysis using random effect models suggested that the combined DOR was 10.67 (95% CI = 3.73–30.53). In addition, the threshold effect analysis showed that the threshold effect was 0.23 and the correlation coefficient was −0.48, indicating that there was no threshold effect. In the forest map, the DOR of each study and the combined DOR are not distributed along the same line, and Q = 2.2 × 1014, P ≤ 0.001. Furthermore, the metaregression analysis of PCT study design, bacterial infection site, and mean age displayed that the P values were >0.05. The combined sensitivity was 0.77 (95% CI = 0.54–0.90), the combined specificity was 0.76 (95% CI = 0.70–0.82), the combined positive likelihood ratio was 3.25 (95% CI = 2.33–4.52), the combined negative likelihood ratio was 0.30 (95% CI = 0.14–0.67), and the combined AUC was 0.80 (95% CI = 0.76–0.83). In conclusion, PCT has moderate diagnostic value for adult liver failure complicated with bacterial infection, and it is a better auxiliary diagnostic index for liver failure with bacterial infection. However, the results of procalcitonin must be carefully interpreted combined with medical history, physical examination, and microbiological assessment.