R. Hernández-Clemente, A. Hornero, V. González-Dugo, M. Berdugo, J. Quero, J. Jiménez, F. Maestre
{"title":"Global monitoring of soil multifunctionality in drylands using satellite imagery and field data","authors":"R. Hernández-Clemente, A. Hornero, V. González-Dugo, M. Berdugo, J. Quero, J. Jiménez, F. Maestre","doi":"10.1002/rse2.340","DOIUrl":"https://doi.org/10.1002/rse2.340","url":null,"abstract":"","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"45709191","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
James G. C. Ball, Sebastian H. M. Hickman, Tobias D. Jackson, Xian Jing Koay, James Hirst, William Jay, Matthew Archer, Mélaine Aubry‐Kientz, Grégoire Vincent, David A. Coomes
Abstract Tropical forests are a major component of the global carbon cycle and home to two‐thirds of terrestrial species. Upper‐canopy trees store the majority of forest carbon and can be vulnerable to drought events and storms. Monitoring their growth and mortality is essential to understanding forest resilience to climate change, but in the context of forest carbon storage, large trees are underrepresented in traditional field surveys, so estimates are poorly constrained. Aerial photographs provide spectral and textural information to discriminate between tree crowns in diverse, complex tropical canopies, potentially opening the door to landscape monitoring of large trees. Here we describe a new deep convolutional neural network method, Detectree2 , which builds on the Mask R‐CNN computer vision framework to recognize the irregular edges of individual tree crowns from airborne RGB imagery. We trained and evaluated this model with 3797 manually delineated tree crowns at three sites in Malaysian Borneo and one site in French Guiana. As an example application, we combined the delineations with repeat lidar surveys (taken between 3 and 6 years apart) of the four sites to estimate the growth and mortality of upper‐canopy trees. Detectree2 delineated 65 000 upper‐canopy trees across 14 km 2 of aerial images. The skill of the automatic method in delineating unseen test trees was good ( F 1 score = 0.64) and for the tallest category of trees was excellent ( F 1 score = 0.74). As predicted from previous field studies, we found that growth rate declined with tree height and tall trees had higher mortality rates than intermediate‐size trees. Our approach demonstrates that deep learning methods can automatically segment trees in widely accessible RGB imagery. This tool (provided as an open‐source Python package) has many potential applications in forest ecology and conservation, from estimating carbon stocks to monitoring forest phenology and restoration. Python package available to install at https://github.com/PatBall1/Detectree2 .
热带森林是全球碳循环的主要组成部分，也是三分之二陆生物种的家园。上冠层树木储存了大部分的森林碳，可能容易受到干旱事件和风暴的影响。监测它们的生长和死亡对于了解森林对气候变化的适应能力至关重要，但在森林碳储量的背景下，传统的实地调查中大树的代表性不足，因此估算结果的约束很差。航空照片提供了光谱和纹理信息，可以区分不同、复杂的热带树冠中的树冠，这可能为大型树木的景观监测打开了大门。在这里，我们描述了一种新的深度卷积神经网络方法Detectree2，它建立在Mask R - CNN计算机视觉框架的基础上，从机载RGB图像中识别单个树冠的不规则边缘。我们在马来西亚婆罗洲的三个地点和法属圭亚那的一个地点对3797个人工绘制的树冠进行了训练和评估。作为一个应用实例，我们结合了四个地点的重复激光雷达调查(间隔3到6年)来估计上冠层树木的生长和死亡率。Detectree2在14公里的航空图像中描绘了65000棵上冠层树木。自动方法对未见测试树的圈定能力较好(f1得分= 0.64)，对最高类别树的圈定能力较好(f1得分= 0.74)。正如以前的野外研究预测的那样，我们发现生长速率随树高而下降，高大树木的死亡率高于中等大小的树木。我们的方法表明，深度学习方法可以在广泛访问的RGB图像中自动分割树。这个工具(作为一个开源的Python包提供)在森林生态和保护中有许多潜在的应用，从估算碳储量到监测森林物候和恢复。Python包可在https://github.com/PatBall1/Detectree2上安装。
{"title":"Accurate delineation of individual tree crowns in tropical forests from aerial <scp>RGB</scp> imagery using Mask <scp>R‐CNN</scp>","authors":"James G. C. Ball, Sebastian H. M. Hickman, Tobias D. Jackson, Xian Jing Koay, James Hirst, William Jay, Matthew Archer, Mélaine Aubry‐Kientz, Grégoire Vincent, David A. Coomes","doi":"10.1002/rse2.332","DOIUrl":"https://doi.org/10.1002/rse2.332","url":null,"abstract":"Abstract Tropical forests are a major component of the global carbon cycle and home to two‐thirds of terrestrial species. Upper‐canopy trees store the majority of forest carbon and can be vulnerable to drought events and storms. Monitoring their growth and mortality is essential to understanding forest resilience to climate change, but in the context of forest carbon storage, large trees are underrepresented in traditional field surveys, so estimates are poorly constrained. Aerial photographs provide spectral and textural information to discriminate between tree crowns in diverse, complex tropical canopies, potentially opening the door to landscape monitoring of large trees. Here we describe a new deep convolutional neural network method, Detectree2 , which builds on the Mask R‐CNN computer vision framework to recognize the irregular edges of individual tree crowns from airborne RGB imagery. We trained and evaluated this model with 3797 manually delineated tree crowns at three sites in Malaysian Borneo and one site in French Guiana. As an example application, we combined the delineations with repeat lidar surveys (taken between 3 and 6 years apart) of the four sites to estimate the growth and mortality of upper‐canopy trees. Detectree2 delineated 65 000 upper‐canopy trees across 14 km 2 of aerial images. The skill of the automatic method in delineating unseen test trees was good ( F 1 score = 0.64) and for the tallest category of trees was excellent ( F 1 score = 0.74). As predicted from previous field studies, we found that growth rate declined with tree height and tall trees had higher mortality rates than intermediate‐size trees. Our approach demonstrates that deep learning methods can automatically segment trees in widely accessible RGB imagery. This tool (provided as an open‐source Python package) has many potential applications in forest ecology and conservation, from estimating carbon stocks to monitoring forest phenology and restoration. Python package available to install at https://github.com/PatBall1/Detectree2 .","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":"222 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-05-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135239414","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ben. G. Weinstein, S. Marconi, Sarah J. Graves, Alina Zare, Aditya Singh, Stephanie A. Bohlman, L. Magee, Daniel J. Johnson, P. Townsend, E. White
Measuring forest biodiversity using terrestrial surveys is expensive and can only capture common species abundance in large heterogeneous landscapes. In contrast, combining airborne imagery with computer vision can generate individual tree data at the scales of hundreds of thousands of trees. To train computer vision models, ground‐based species labels are combined with airborne reflectance data. Due to the difficulty of finding rare species in a large landscape, many classification models only include the most abundant species, leading to biased predictions at broad scales. For example, if only common species are used to train the model, this assumes that these samples are representative across the entire landscape. Extending classification models to include rare species requires targeted data collection and algorithmic improvements to overcome large data imbalances between dominant and rare taxa. We use a targeted sampling workflow to the Ordway Swisher Biological Station within the US National Ecological Observatory Network (NEON), where traditional forestry plots had identified six canopy tree species with more than 10 individuals at the site. Combining iterative model development with rare species sampling, we extend a training dataset to include 14 species. Using a multi‐temporal hierarchical model, we demonstrate the ability to include species predicted at <1% frequency in landscape without losing performance on the dominant species. The final model has over 75% accuracy for 14 species with improved rare species classification compared to 61% accuracy of a baseline deep learning model. After filtering out dead trees, we generate landscape species maps of individual crowns for over 670 000 individual trees. We find distinct patches of forest composed of rarer species at the full‐site scale, highlighting the importance of capturing species diversity in training data. We estimate the relative abundance of 14 species within the landscape and provide three measures of uncertainty to generate a range of counts for each species. For example, we estimate that the dominant species, Pinus palustris accounts for c. 28% of predicted stems, with models predicting a range of counts between 160 000 and 210 000 individuals. These maps provide the first estimates of canopy tree diversity within a NEON site to include rare species and provide a blueprint for capturing tree diversity using airborne computer vision at broad scales.
{"title":"Capturing long‐tailed individual tree diversity using an airborne imaging and a multi‐temporal hierarchical model","authors":"Ben. G. Weinstein, S. Marconi, Sarah J. Graves, Alina Zare, Aditya Singh, Stephanie A. Bohlman, L. Magee, Daniel J. Johnson, P. Townsend, E. White","doi":"10.1002/rse2.335","DOIUrl":"https://doi.org/10.1002/rse2.335","url":null,"abstract":"Measuring forest biodiversity using terrestrial surveys is expensive and can only capture common species abundance in large heterogeneous landscapes. In contrast, combining airborne imagery with computer vision can generate individual tree data at the scales of hundreds of thousands of trees. To train computer vision models, ground‐based species labels are combined with airborne reflectance data. Due to the difficulty of finding rare species in a large landscape, many classification models only include the most abundant species, leading to biased predictions at broad scales. For example, if only common species are used to train the model, this assumes that these samples are representative across the entire landscape. Extending classification models to include rare species requires targeted data collection and algorithmic improvements to overcome large data imbalances between dominant and rare taxa. We use a targeted sampling workflow to the Ordway Swisher Biological Station within the US National Ecological Observatory Network (NEON), where traditional forestry plots had identified six canopy tree species with more than 10 individuals at the site. Combining iterative model development with rare species sampling, we extend a training dataset to include 14 species. Using a multi‐temporal hierarchical model, we demonstrate the ability to include species predicted at <1% frequency in landscape without losing performance on the dominant species. The final model has over 75% accuracy for 14 species with improved rare species classification compared to 61% accuracy of a baseline deep learning model. After filtering out dead trees, we generate landscape species maps of individual crowns for over 670 000 individual trees. We find distinct patches of forest composed of rarer species at the full‐site scale, highlighting the importance of capturing species diversity in training data. We estimate the relative abundance of 14 species within the landscape and provide three measures of uncertainty to generate a range of counts for each species. For example, we estimate that the dominant species, Pinus palustris accounts for c. 28% of predicted stems, with models predicting a range of counts between 160 000 and 210 000 individuals. These maps provide the first estimates of canopy tree diversity within a NEON site to include rare species and provide a blueprint for capturing tree diversity using airborne computer vision at broad scales.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42181383","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
G. Krivek, Alexander Gillert, Martin Harder, M. Fritze, Karina Frankowski, Luisa Timm, Liska Meyer‐Olbersleben, Uwe Freiherr von Lukas, G. Kerth, J. van Schaik
{"title":"BatNet\u0000 : a deep learning‐based tool for automated bat species identification from camera trap images","authors":"G. Krivek, Alexander Gillert, Martin Harder, M. Fritze, Karina Frankowski, Luisa Timm, Liska Meyer‐Olbersleben, Uwe Freiherr von Lukas, G. Kerth, J. van Schaik","doi":"10.1002/rse2.339","DOIUrl":"https://doi.org/10.1002/rse2.339","url":null,"abstract":"","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"43714566","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
M. Villoslada, H. Ylänne, S. Juutinen, T. Kolari, Pasi Korpelainen, T. Tahvanainen, Franziska Wolff, T. Kumpula
Herbivores can exert a controlling effect on the reproduction and growth of shrubs, thereby counter‐acting the climate‐driven encroachment of shrubs in the Arctic and the potential consequences. This control is particularly evident in the case of abundant herbivores, such as reindeer (Rangifer tarandus tarandus), whose grazing patterns are affected by management. Here, we tested how different reindeer grazing practices on the border between Finland and Norway impact the occurrence of willow (Salix spp.) dominated patches, their above‐ground biomass (AGB) and the ability of willows to form dense thickets. We used a combination of multispectral and RGB imagery obtained from unoccupied aerial vehicles field data and an ensemble of machine‐learning models, which allowed us to model the occurrence of plant community types (Overall accuracy = 0.80), AGB fractions (maximum R2 = 0.90) and topsoil moisture (maximum R2 = 0.89). With this combination of approaches, we show that willows are kept in a browsing‐trap under spring and early summer grazing by reindeer, growing mostly small and scattered in the landscape. In contrast, willows under the winter grazing regime formed dense stands, particularly within riparian areas. We confirm this pattern using a random forest willow habitat distribution model based on topographical parameters. The model shows that willow biomass correlated with parameters of optimal habitat quality only in the winter grazing regime and did not respond to the same parameters under spring and summer grazing of reindeer.
{"title":"Reindeer control over shrubification in subarctic wetlands: spatial analysis based on unoccupied aerial vehicle imagery","authors":"M. Villoslada, H. Ylänne, S. Juutinen, T. Kolari, Pasi Korpelainen, T. Tahvanainen, Franziska Wolff, T. Kumpula","doi":"10.1002/rse2.337","DOIUrl":"https://doi.org/10.1002/rse2.337","url":null,"abstract":"Herbivores can exert a controlling effect on the reproduction and growth of shrubs, thereby counter‐acting the climate‐driven encroachment of shrubs in the Arctic and the potential consequences. This control is particularly evident in the case of abundant herbivores, such as reindeer (Rangifer tarandus tarandus), whose grazing patterns are affected by management. Here, we tested how different reindeer grazing practices on the border between Finland and Norway impact the occurrence of willow (Salix spp.) dominated patches, their above‐ground biomass (AGB) and the ability of willows to form dense thickets. We used a combination of multispectral and RGB imagery obtained from unoccupied aerial vehicles field data and an ensemble of machine‐learning models, which allowed us to model the occurrence of plant community types (Overall accuracy = 0.80), AGB fractions (maximum R2 = 0.90) and topsoil moisture (maximum R2 = 0.89). With this combination of approaches, we show that willows are kept in a browsing‐trap under spring and early summer grazing by reindeer, growing mostly small and scattered in the landscape. In contrast, willows under the winter grazing regime formed dense stands, particularly within riparian areas. We confirm this pattern using a random forest willow habitat distribution model based on topographical parameters. The model shows that willow biomass correlated with parameters of optimal habitat quality only in the winter grazing regime and did not respond to the same parameters under spring and summer grazing of reindeer.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41812984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Tom E. Verhelst, P. Vangansbeke, P. De Frenne, Barbara D'hont, Q. Ponette, Luc Willems, H. Verbeeck, K. Calders
Forest edges can be important strongholds for biodiversity and play a crucial role in the protection of forest interiors against edge effects. However, their potential to host biodiversity is dependent on the structure of the forest: Abrupt edges often fail to realise this potential. Yet, methods to accurately characterise and quantify forest edge abruptness are currently lacking. Here, we combine three‐dimensional forest structural data with biodiversity monitoring to assess the influence of forest edge structure on habitat suitability. We derived several structural metrics to determine forest edge abruptness using terrestrial laser scanning and applied these to six forest edge transects in Belgium. The local soundscapes were captured using audio recording devices (Audiomoths) and quantified using acoustic indices (AIs) (metrics on the soundscape characteristics). In each transect, the dawn choruses were recorded over a period of a week, both at the edge and the interior of the forest. No correlation between the AIs and bird species richness was found. There were clear differences between transects in the structural metrics and the recorded soundscapes. Some possible relations between both were found. In this proof of concept, we demonstrated innovative techniques to semi‐automatically classify forest structure and rapidly quantify soundscape characteristics and found a weak effect of forest edge structure on bird biophony.
{"title":"Forest edge structure from terrestrial laser scanning to explain bird biophony characteristics from acoustic indices","authors":"Tom E. Verhelst, P. Vangansbeke, P. De Frenne, Barbara D'hont, Q. Ponette, Luc Willems, H. Verbeeck, K. Calders","doi":"10.1002/rse2.334","DOIUrl":"https://doi.org/10.1002/rse2.334","url":null,"abstract":"Forest edges can be important strongholds for biodiversity and play a crucial role in the protection of forest interiors against edge effects. However, their potential to host biodiversity is dependent on the structure of the forest: Abrupt edges often fail to realise this potential. Yet, methods to accurately characterise and quantify forest edge abruptness are currently lacking. Here, we combine three‐dimensional forest structural data with biodiversity monitoring to assess the influence of forest edge structure on habitat suitability. We derived several structural metrics to determine forest edge abruptness using terrestrial laser scanning and applied these to six forest edge transects in Belgium. The local soundscapes were captured using audio recording devices (Audiomoths) and quantified using acoustic indices (AIs) (metrics on the soundscape characteristics). In each transect, the dawn choruses were recorded over a period of a week, both at the edge and the interior of the forest. No correlation between the AIs and bird species richness was found. There were clear differences between transects in the structural metrics and the recorded soundscapes. Some possible relations between both were found. In this proof of concept, we demonstrated innovative techniques to semi‐automatically classify forest structure and rapidly quantify soundscape characteristics and found a weak effect of forest edge structure on bird biophony.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"41886742","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jiale Chen, Li Wang, T. Jucker, Hongzhi Da, Zhaochen Zhang, Jianbo Hu, Qingsong Yang, Xihua Wang, Yuchu Qin, Guochun Shen, Li Shu, Jian Zhang
Accurate and efficient detection of canopy gaps is essential for understanding species regeneration and community dynamics in forests. Unoccupied aerial vehicles (UAVs) equipped with visible light (e.g., RGB) cameras have the potential to be one of the most cost‐effective approaches for detecting gaps. However, current gap‐detection methods based on spectral, textural, and/or structural information derived from UAV RGB imagery are unreliable in species‐rich forests with complex terrain due to high spectral complexity and topographic shadowing. Here, we compared the performance of four methods, including pixel‐based supervised classification (PBSC), object‐based classification (OBIA), Canopy Height Model thresholding classification, and HSTAC [a novel method we developed which combines Photographic Height (H), Spectral (S), and Textural (T) information for Automatic Classification (AC)] for characterizing canopy gaps in a 20‐ha permanent subtropical forest plot of eastern China. All classification results were evaluated through a comparison with canopy gaps detected from both field surveys and UAV‐borne LiDAR data. Among the four classification methods, HSTAC performed best in terms of detection efficiency (96% overall accuracy when compared to field data and 85% when compared to the LiDAR data), classification accuracy (3–18% improvement compared to alternative methods), and speed (1–1.5 h faster on the same machine). Of the four topographic factors (elevation, slope, aspect, and convexity), elevation was the one that most affected the accuracy of canopy gap detection. The errors of PBSC classification mainly came from the gaps at low elevations, while OBIA located the position of gaps well but overestimated their sizes. Overall, HSTAC avoids many of the inherent limitations of current state‐of‐the‐art methods and can accurately map canopy gaps in diverse subtropical forests with complex terrain. Our study provides a suitable way for long‐term forest canopy monitoring, real‐time applications, and contributes to a better understanding of forest plant community assembly and succession dynamics.
{"title":"Detecting forest canopy gaps using unoccupied aerial vehicle\u0000 RGB\u0000 imagery in a species‐rich subtropical forest","authors":"Jiale Chen, Li Wang, T. Jucker, Hongzhi Da, Zhaochen Zhang, Jianbo Hu, Qingsong Yang, Xihua Wang, Yuchu Qin, Guochun Shen, Li Shu, Jian Zhang","doi":"10.1002/rse2.336","DOIUrl":"https://doi.org/10.1002/rse2.336","url":null,"abstract":"Accurate and efficient detection of canopy gaps is essential for understanding species regeneration and community dynamics in forests. Unoccupied aerial vehicles (UAVs) equipped with visible light (e.g., RGB) cameras have the potential to be one of the most cost‐effective approaches for detecting gaps. However, current gap‐detection methods based on spectral, textural, and/or structural information derived from UAV RGB imagery are unreliable in species‐rich forests with complex terrain due to high spectral complexity and topographic shadowing. Here, we compared the performance of four methods, including pixel‐based supervised classification (PBSC), object‐based classification (OBIA), Canopy Height Model thresholding classification, and HSTAC [a novel method we developed which combines Photographic Height (H), Spectral (S), and Textural (T) information for Automatic Classification (AC)] for characterizing canopy gaps in a 20‐ha permanent subtropical forest plot of eastern China. All classification results were evaluated through a comparison with canopy gaps detected from both field surveys and UAV‐borne LiDAR data. Among the four classification methods, HSTAC performed best in terms of detection efficiency (96% overall accuracy when compared to field data and 85% when compared to the LiDAR data), classification accuracy (3–18% improvement compared to alternative methods), and speed (1–1.5 h faster on the same machine). Of the four topographic factors (elevation, slope, aspect, and convexity), elevation was the one that most affected the accuracy of canopy gap detection. The errors of PBSC classification mainly came from the gaps at low elevations, while OBIA located the position of gaps well but overestimated their sizes. Overall, HSTAC avoids many of the inherent limitations of current state‐of‐the‐art methods and can accurately map canopy gaps in diverse subtropical forests with complex terrain. Our study provides a suitable way for long‐term forest canopy monitoring, real‐time applications, and contributes to a better understanding of forest plant community assembly and succession dynamics.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"48966663","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Na Chen, N. Tsendbazar, Daniela Requena Suarez, J. Verbesselt, M. Herold
Characterization of regrowing forests is vital for understanding forest dynamics to assess the impacts on carbon stocks and to support sustainable forest management. Although remote sensing is a key tool for understanding and monitoring forest dynamics, the use of exclusively remotely sensed data to explore the effects of different variables on regrowing forests across all biomes in Brazil has rarely been investigated. Here, we analyzed how environmental and human factors affect regrowing forests. Based on Brazil's secondary forest age map, 3060 locations disturbed between 1984 and 2018 were sampled, interpreted and analyzed in different biomes. We interpreted the time since disturbance for the sampled pixels in Google Earth Engine. Elevation, slope, climatic water deficit (CWD), the total Nitrogen of soil, cation exchange capacity (CEC) of soil, surrounding tree cover, distance to roads, distance to settlements and fire frequency were analyzed in their importance for predicting aboveground biomass (AGB) and tree cover derived from global forest aboveground biomass map and tree cover map, respectively. Results show that time since disturbance interpreted from satellite time series is the most important predictor for characterizing AGB and tree cover of regrowing forests. AGB increased with increasing time since disturbance, surrounding tree cover, soil total N, slope, distance to roads, distance to settlements and decreased with larger fire frequency, CWD and CEC of soil. Tree cover increased with larger time since disturbance, soil total N, surrounding tree cover, distance to roads, distance to settlements, slope and decreased with increasing elevation and CWD. These results emphasize the importance of remotely sensing products as key opportunities to improve the characterization of forest regrowth and to reduce data gaps and uncertainties related to forest carbon sink estimation. Our results provide a better understanding of regional forest dynamics, toward developing and assessing effective forest‐related restoration and climatic mitigation strategies.
{"title":"Characterizing aboveground biomass and tree cover of regrowing forests in Brazil using multi‐source remote sensing data","authors":"Na Chen, N. Tsendbazar, Daniela Requena Suarez, J. Verbesselt, M. Herold","doi":"10.1002/rse2.328","DOIUrl":"https://doi.org/10.1002/rse2.328","url":null,"abstract":"Characterization of regrowing forests is vital for understanding forest dynamics to assess the impacts on carbon stocks and to support sustainable forest management. Although remote sensing is a key tool for understanding and monitoring forest dynamics, the use of exclusively remotely sensed data to explore the effects of different variables on regrowing forests across all biomes in Brazil has rarely been investigated. Here, we analyzed how environmental and human factors affect regrowing forests. Based on Brazil's secondary forest age map, 3060 locations disturbed between 1984 and 2018 were sampled, interpreted and analyzed in different biomes. We interpreted the time since disturbance for the sampled pixels in Google Earth Engine. Elevation, slope, climatic water deficit (CWD), the total Nitrogen of soil, cation exchange capacity (CEC) of soil, surrounding tree cover, distance to roads, distance to settlements and fire frequency were analyzed in their importance for predicting aboveground biomass (AGB) and tree cover derived from global forest aboveground biomass map and tree cover map, respectively. Results show that time since disturbance interpreted from satellite time series is the most important predictor for characterizing AGB and tree cover of regrowing forests. AGB increased with increasing time since disturbance, surrounding tree cover, soil total N, slope, distance to roads, distance to settlements and decreased with larger fire frequency, CWD and CEC of soil. Tree cover increased with larger time since disturbance, soil total N, surrounding tree cover, distance to roads, distance to settlements, slope and decreased with increasing elevation and CWD. These results emphasize the importance of remotely sensing products as key opportunities to improve the characterization of forest regrowth and to reduce data gaps and uncertainties related to forest carbon sink estimation. Our results provide a better understanding of regional forest dynamics, toward developing and assessing effective forest‐related restoration and climatic mitigation strategies.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-04-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42443337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zhongqi Miao, Stella X. Yu, K. Landolt, M. Koneff, Timothy P. White, Luke J. Fara, E. Hlavacek, B. Pickens, Travis J. Harrison, W. Getz
Remote aerial sensing provides a non‐invasive, large geographical‐scale technology for avian monitoring, but the manual processing of images limits its development and applications. Artificial Intelligence (AI) methods can be used to mitigate this manual image processing requirement. The implementation of AI methods, however, has several challenges: (1) imbalanced (i.e., long‐tailed) data distribution, (2) annotation uncertainty in categorization, and (3) dataset discrepancies across different study sites. Here we use aerial imagery data of waterbirds around Cape Cod and Lake Michigan in the United States to examine how these challenges limit avian recognition performance. We review existing solutions and demonstrate as use cases how methods like Label Distribution Aware Marginal Loss with Deferred Re‐Weighting, hierarchical classification, and FixMatch address the three challenges. We also present a new approach to tackle the annotation uncertainty challenge using a Soft‐fine Pseudo‐Label methodology. Finally, we aim with this paper to increase awareness in the ecological remote sensing community of these challenges and bridge the gap between ecological applications and state‐of‐the‐art computer science, thereby opening new doors to future research.
{"title":"Challenges and solutions for automated avian recognition in aerial imagery","authors":"Zhongqi Miao, Stella X. Yu, K. Landolt, M. Koneff, Timothy P. White, Luke J. Fara, E. Hlavacek, B. Pickens, Travis J. Harrison, W. Getz","doi":"10.1002/rse2.318","DOIUrl":"https://doi.org/10.1002/rse2.318","url":null,"abstract":"Remote aerial sensing provides a non‐invasive, large geographical‐scale technology for avian monitoring, but the manual processing of images limits its development and applications. Artificial Intelligence (AI) methods can be used to mitigate this manual image processing requirement. The implementation of AI methods, however, has several challenges: (1) imbalanced (i.e., long‐tailed) data distribution, (2) annotation uncertainty in categorization, and (3) dataset discrepancies across different study sites. Here we use aerial imagery data of waterbirds around Cape Cod and Lake Michigan in the United States to examine how these challenges limit avian recognition performance. We review existing solutions and demonstrate as use cases how methods like Label Distribution Aware Marginal Loss with Deferred Re‐Weighting, hierarchical classification, and FixMatch address the three challenges. We also present a new approach to tackle the annotation uncertainty challenge using a Soft‐fine Pseudo‐Label methodology. Finally, we aim with this paper to increase awareness in the ecological remote sensing community of these challenges and bridge the gap between ecological applications and state‐of‐the‐art computer science, thereby opening new doors to future research.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-04-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"42799043","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Climate change modifies plant phenology through shifts in seasonal temperature and precipitation. Because the timing of plant growth can limit herbivore population dynamics, climatic alteration of historical patterns of vegetation seasonality may alter population trajectories in such taxa. Thus, sound management decisions may depend on understanding how plant growth varies across a landscape within and among distinct management units or protected areas. Here, we examine spatial variation in the timing of spring plant growth, measured using a network of automated time‐lapse cameras distributed across the range of endangered Sierra Nevada bighorn sheep (Ovis canadensis sierrae) in California, USA. We tracked greenness of individual plants across 2 years to compare spatial patterns of forage phenology in snowy and drought years. Green‐up timing was derived for individual plants across the camera network and compared with local estimates of green‐up timing from satellite data. Satellite‐derived estimates of green‐up timing showed strong correspondence with camera‐derived estimates in areas with dense vegetation cover and weak correspondence in areas with sparse vegetation cover. Daily time‐lapse imagery revealed consistent variation in green‐up timing across elevation, both among latitudinal zones and among individual plant species. Green‐up timing was earlier in 2020 than in 2019, reflecting differences in the end of the snowy season. Because bighorn forage seasonally on alpine species with a brief growing period, spring migration of bighorn may be linked to variation in snowmelt and plant growth across elevational gradients.
{"title":"Fine‐scale landscape phenology revealed through time‐lapse imagery: implications for conservation and management of an endangered migratory herbivore","authors":"C. John, Jeffrey T. Kerby, T. Stephenson, E. Post","doi":"10.1002/rse2.331","DOIUrl":"https://doi.org/10.1002/rse2.331","url":null,"abstract":"Climate change modifies plant phenology through shifts in seasonal temperature and precipitation. Because the timing of plant growth can limit herbivore population dynamics, climatic alteration of historical patterns of vegetation seasonality may alter population trajectories in such taxa. Thus, sound management decisions may depend on understanding how plant growth varies across a landscape within and among distinct management units or protected areas. Here, we examine spatial variation in the timing of spring plant growth, measured using a network of automated time‐lapse cameras distributed across the range of endangered Sierra Nevada bighorn sheep (Ovis canadensis sierrae) in California, USA. We tracked greenness of individual plants across 2 years to compare spatial patterns of forage phenology in snowy and drought years. Green‐up timing was derived for individual plants across the camera network and compared with local estimates of green‐up timing from satellite data. Satellite‐derived estimates of green‐up timing showed strong correspondence with camera‐derived estimates in areas with dense vegetation cover and weak correspondence in areas with sparse vegetation cover. Daily time‐lapse imagery revealed consistent variation in green‐up timing across elevation, both among latitudinal zones and among individual plant species. Green‐up timing was earlier in 2020 than in 2019, reflecting differences in the end of the snowy season. Because bighorn forage seasonally on alpine species with a brief growing period, spring migration of bighorn may be linked to variation in snowmelt and plant growth across elevational gradients.","PeriodicalId":21132,"journal":{"name":"Remote Sensing in Ecology and Conservation","volume":" ","pages":""},"PeriodicalIF":5.5,"publicationDate":"2023-04-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44936886","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：