Article
Geochemistry & Geophysics
Ming Liu, Zhongqiu Sun, Shan Lu, Kenji Omasa
Summary: The combination of multiangular, polarimetric, and hyperspectral measurements improves the accuracy of estimating leaf nitrogen concentration and enhances the effectiveness of hyperspectral indices across a wide range of viewing angles. Additionally, polarimetric measurements deepen our understanding of optical properties of light reflected from leaves.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Environmental Sciences
Liang Wan, Weijun Zhou, Yong He, Thomas Cherico Wanger, Haiyan Cen
Summary: In this study, we propose a new transfer learning method called TCA-SVR to transfer leaf nitrogen concentration (LNC) assessment models across different plant species. By analyzing five remote sensing datasets, we find that combining visible, near infrared, and shortwave infrared reflectance achieves the optimal LNC assessment across all datasets. Compared to the established PLSR model, TCA-SVR greatly improves the transferability of the LNC assessment model and further improves performance through model updating.
REMOTE SENSING OF ENVIRONMENT
(2022)
Article
Agriculture, Multidisciplinary
Jingang Wang, Tian Tian, Haijiang Wang, Jing Cui, Yongqi Zhu, Wenxu Zhang, Xuanmeng Tong, Tianhang Zhou, Zhenkang Yang, Jiaqi Sun
Summary: The model based on the combination of LNC and OA-sensitive bands improves the accuracy and universality of cotton leaf nitrogen concentration estimation. The PLSR models show higher accuracy and stability compared to PCR models.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Environmental Sciences
Cao Dinh Dung, Stephen J. Trueman, Helen M. Wallace, Michael B. Farrar, Tsvakai Gama, Iman Tahmasbian, Shahla Hosseini Bai
Summary: This study evaluated the potential of hyperspectral imaging to estimate nutrient concentrations and predict yield in strawberry plants. The results showed that the prediction accuracy was higher for leaves, flowers, and unripe fruit compared to ripe fruit.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2023)
Article
Agriculture, Multidisciplinary
Heli Li, Guijun Yang, Huiling Long, Haikuan Feng, Bo Xu, Chunjiang Zhao
Summary: The characteristic coefficient of vertical leaf nitrogen (N) profile is a crucial parameter for crop management. In this study, we propose a robust model to estimate this coefficient using canopy spectral reflectance. We evaluate and compare the accuracy and stability of models using various statistical measures. The results demonstrate that the proposed model accurately estimates the coefficient, enabling nondestructive assessment and tracking of large-scale leaf nitrogen dynamics.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2023)
Article
Environmental Sciences
Caixia Yin, Xin Lv, Lifu Zhang, Lulu Ma, Huihan Wang, Linshan Zhang, Ze Zhang
Summary: This study tested the feasibility of using a UAV equipped with a hyperspectral spectrometer to monitor cotton leaf nitrogen content. The results showed that collecting UAV hyperspectral images at multiple heights improved the accuracy of assessing cotton nitrogen content.
Article
Ecology
Guangman Song, Quan Wang
Summary: Information on plant species is crucial for forest ecosystems, including biodiversity monitoring and forest management. Traditional plant species inventory methods are inefficient in terms of cost and performance, necessitating the development of a quick and feasible approach. Remote sensing has emerged as an active approach for plant species classification, but only a few studies have utilized hyperspectral information. This study effectively classifies plant species using hyperspectral leaf information and machine learning models, optimized through Bayesian optimization, leading to improved classification accuracy.
ECOLOGICAL INFORMATICS
(2023)
Article
Environmental Sciences
Kaiyi Bi, Zheng Niu, Shunfu Xiao, Jie Bai, Gang Sun, Ji Wang, Zeying Han, Shuai Gao
Summary: This study tested the ability of Hyperspectral LiDAR (HSL) in estimating maize nitrogen concentration and biomass, showing that Partial Least Squares Regression (PLSR) performed better with HSL assistance. Future research should utilize larger datasets to test the viability of using HSL for monitoring crop nitrogen concentration.
Article
Agriculture, Multidisciplinary
Ziling Chen, Jialei Wang, Tao Wang, Zhihang Song, Yikai Li, Yuanmeng Huang, Liangju Wang, Jian Jin
Summary: Hyperspectral Imaging (HSI) is widely used in field plant phenotyping, but current imaging systems have limitations. A robotic system can replace human operators to improve efficiency and accuracy in hyperspectral imaging.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2021)
Article
Remote Sensing
Haibo Yang, Fei Li, Yuncai Hu, Kang Yu
Summary: Many empirical models based on hyperspectral indices have been developed to estimate nitrogen status of crops, with a focus on sensitive bands identification. However, this study found that band optimization and formula formats are crucial for achieving the best performance of these indices. The optimized HIs showed more robust performances for canopy N concentration prediction compared to published indices, with band optimization significantly improving performance by 16%-71% and formula formats affecting explanatory power by 3%-18%. The results highlight the potential of hyperspectral sensing for improving the estimation of field crops CNC.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2021)
Article
Plant Sciences
Armando Sterling, Julio A. Di Rienzo
Summary: This study aimed to detect and classify South American Leaf Blight (SALB) in rubber trees using visible and near-infrared sensors as well as predictive modeling. The results showed that the random forest, artificial neural network, and bagged classification and regression trees models had excellent classification performance and could accurately predict the levels of SALB. However, the prediction performance for photochemical traits was lower.
Article
Horticulture
Renan Tosin, Isabel Pocas, Helena Novo, Jorge Teixeira, Natacha Fontes, Antonio Graca, Mario Cunha
Summary: This study developed two models to estimate Psi(pd) in a commercial vineyard, utilizing spectral data and machine learning algorithms. The first model estimated Psi(pd) based on vine canopy reflectance and selected suitable vegetation indices, while the second model optimized variables for Psi(pd) estimation based on pigments' concentrations assessed through hyperspectral reflectance. The B-MARS algorithm produced the best results with a RRMSE between 13-14% in validation.
SCIENTIA HORTICULTURAE
(2021)
Article
Agronomy
Xiaoyu Zhi, Sean Reynolds Massey-Reed, Alex Wu, Andries Potgieter, Andrew Borrell, Colleen Hunt, David Jordan, Yan Zhao, Scott Chapman, Graeme Hammer, Barbara George-Jaeggli
Summary: This study used sorghum as a model to predict photosynthetic capacity traits using hyperspectral sensing and genetic analysis. The researchers identified candidate genes associated with these traits and demonstrated the potential of this method for screening large germplasm collections for enhanced photosynthesis.
Article
Environmental Sciences
Yelu Zeng, Dalei Hao, Grayson Badgley, Alexander Damm, Uwe Rascher, Youngryel Ryu, Jennifer Johnson, Vera Krieger, Shengbiao Wu, Han Qiu, Yaling Liu, Joseph A. Berry, Min Chen
Summary: Disentangling the individual contributions from vegetation and soil in measured canopy reflectance is difficult. Solar Induced chlorophyll Fluorescence (SIF) can help separate vegetation and soil components, with NIRvH showing the smallest offset compared to NIRv and DVI in isolating true NIR reflectance of vegetation. This study highlights the potential of NIRvH in retrieving canopy structure parameters and estimating fluorescence yield using hyperspectral measurements.
REMOTE SENSING OF ENVIRONMENT
(2021)
Article
Multidisciplinary Sciences
Aldemar Reyes-Trujillo, Martha C. Daza-Torres, Carlos A. Galindez-Jamioy, Esteban E. Rosero-Garcia, Fernando Munoz-Arboleda, Efrain Solarte-Rodriguez
Summary: The study aimed to develop estimation models to explain nitrogen variations over time based on three spectral data transformations in two growth stages with different nitrogen application levels. The models were built using partial least squares regression analysis trained by three different spectral data transformations.
Article
Computer Science, Information Systems
Yizhuo Li, Teng Fei, Yingjing Huang, Jun Li, Xiang Li, Fan Zhang, Yuhao Kang, Guofeng Wu
Summary: This study proposes a methodological framework for mapping the global geographic distribution of human emotion, utilizing affective computing technology to extract emotions from facial expressions in Flickr photos. The framework combines a species distribution model with physical environment factors and explores different geographic distributions of seven dimensional emotions. The results confirm the effectiveness of the framework and provide new insights into the relationship between human emotions and the physical environment.
INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE
(2021)
Article
Environmental Sciences
Yang Chao, Liu Huizeng, Li Qingquan, Cui Aihong, Xia Rongling, Shi Tiezhu, Zhang Jie, Gao Wenxiu, Zhou Xiang, Wu Guofeng
Summary: The study showed that the urban growth area in the Guangdong-Hong Kong-Macao Greater Bay Area obtained 14.86% from forest loss from 1987 to 2017. The forest loss area in the GBA reached 4040.6 km(2), with 25.60% converted to urban land. Forest loss percentages in Dongguan (19.14%), Guangzhou (18.35%) and Shenzhen (15.81%) were higher than in other cities. Fragmentation of forests increased from 1987-2007, then decreased from 2007-2017. Some forest loss to urban regions shifted from low to high elevation and steep-slope terrains over time, especially in Shenzhen and Hong Kong.
CHINESE GEOGRAPHICAL SCIENCE
(2021)
Article
Forestry
Yuzhi Tang, Quanqin Shao, Tiezhu Shi, Guofeng Wu
Summary: This study investigated the stand volume growth of dominant tree species in a subtropical karst area in Guizhou, and developed growth models based on environmental factors. It revealed that climatic factors and site factors significantly influence stand volume, with topsoil thickness and site quality degree having the strongest positive effect. The findings offer updated knowledge on environmental effects on stand volume growth in subtropical forests in karst areas, and the developed models are useful for forest management and planning, contributing to forest carbon storage assessments and global carbon cycling studies.
Article
Environmental Sciences
Tiezhu Shi, Chao Yang, Huizeng Liu, Chao Wu, Zhihua Wang, He Li, Huifang Zhang, Long Guo, Guofeng Wu, Fenzhen Su
Summary: Due to rapid urbanization in China, lead continues to accumulate in urban topsoil, increasing public exposure risk. This study developed spatial models using proximal and remote sensing data to map lead concentrations in urban topsoil. By extracting landscape factors from the data, they were able to predict soil lead concentrations effectively, with geographically weighted regression achieving better results than regression kriging.
ENVIRONMENTAL POLLUTION
(2021)
Article
Geography, Physical
Jie Zhang, Xuecao Li, Chenchen Zhang, Le Yu, Jingzhe Wang, Xiangyin Wu, Zhongwen Hu, Zihan Zhai, Qingquan Li, Guofeng Wu, Tiezhu Shi
Summary: Rapid economic development and human interference in rapidly urbanized regions have caused significant land use/land cover change (LUCC), impacting ecosystem functions and services. This study examined and predicted the impact of LUCC on ecosystem service value (ESV) in the Guangdong-Hong Kong-Macao Greater Bay Area (GBA) in China from 1990 to 2030. The results showed that the expansion of built-up land had the clearest process during 1990-2015, leading to a decrease in total ESV. The study also demonstrated the effectiveness of the future land use simulation (FLUS) model in predicting future LUCC. The findings emphasize the importance of rational land use and ecological construction for the GBA, providing a reference for ecological planning and environmental protection.
GISCIENCE & REMOTE SENSING
(2022)
Article
Multidisciplinary Sciences
Chao Yang, Huizeng Liu, Qingquan Li, Xuqing Wang, Wei Ma, Cuiling Liu, Xu Fang, Yuzhi Tang, Tiezhu Shi, Qibiao Wang, Yue Xu, Jie Zhang, Xuecao Li, Gang Xu, Junyi Chen, Mo Su, Shuying Wang, Jinjing Wu, Leping Huang, Xue Li, Guofeng Wu
Summary: The study investigates the expansions of human activities in Asian highlands from 2000 to 2020, finding that most of these expansions come from ecological lands. This intensifies habitat fragmentation and results in large ecological costs in low and lower-middle income countries while also supporting Asian developments.
NATURE COMMUNICATIONS
(2022)
Article
Environmental Sciences
Honglin Zhu, Huizeng Liu, Qiming Zhou, Aihong Cui
Summary: This study presents a framework combining machine learning-based downscaling algorithm, residual correction, and precipitation calibration for accurate high-resolution precipitation estimation. The results show that the machine learning-based methods outperform conventional approaches, with spatial random forest and eXtreme gradient boosting performing the best in generating high-resolution precipitation. The geographical difference analysis calibration process significantly improves the downscaled results.
Article
Biodiversity Conservation
Zhongwen Hu, Jinjing Wu, Jingzhe Wang, Yinghui Zhang, Haichao Zhou, Changjun Gao, Junjie Wang, Guofeng Wu
Summary: The study aims to investigate the impact of exotic mangrove species on the spatial dynamics of mangroves in Shenzhen Bay, China. It was found that the mangrove area in the study area increased from 2000 to 2022, with a growth rate of 5.14% in Shenzhen and 2.38% in Hong Kong. The rapid spread of Sonneratia species was identified as one of the main contributors to the growth of mangrove hotspots, mainly concentrated in the estuary delta. The results provide valuable insights for accurate mangrove mapping and emphasize the importance of addressing the spread and invasive potential of exotic mangrove species in the study area, as well as cooperation with adjacent reserves.
ECOLOGICAL INDICATORS
(2023)
Article
Geochemistry & Geophysics
Honglin Zhu, Huizeng Liu, Qiming Zhou, Aihong Cui
Summary: Extreme precipitation events have severe impacts on society, economy, and environment, including floods, flash floods, and landslides. However, the low resolution of satellite-derived precipitation data makes it difficult to quantitatively capture fine-scale heavy rainfall processes. In this study, a downscaling-calibration scheme based on XGBoost algorithm was proposed to improve the spatial resolution and accuracy of satellite-based precipitation extremes. The results showed that the proposed scheme achieved the best performance in reproducing the occurrence and spatial distribution of precipitation, compared to other comparative methods.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Geochemistry & Geophysics
Huizeng Liu, Qingquan Li, Shaopeng Huang, Hong Qiu, Huiping Jiang, Chao Yang, Ping Zhu
Summary: In recent years, novel satellite platforms and sensors have been proposed for the Earth radiation budget (ERB). A study proposed a set of models for estimating the longwave anisotropic factors directly from the Earth's radiative fluxes, aiming to simplify the procedure of simulating the signals of ERB sensors. The models were tested and showed accurate estimation of the anisotropic factors and efficient retrieval of sensor-measured radiances.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Engineering, Electrical & Electronic
Hongxing Cui, Danling Tang, Huizeng Liu, Yi Sui, Xiaowei Gu
Summary: The study proposes a new method based on random forest to predict sea surface height anomaly induced by tropical cyclones. The method utilizes the characteristics of the cyclones and prestorm oceanic parameters as input to accurately predict the anomaly up to 30 days after the cyclone passes through. The proposed method achieves high prediction accuracy in the Western North Pacific region.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Proceedings Paper
Engineering, Aerospace
Ping Zhu, Ozgur Karatekin, Burak Yaglioglu, Huizeng Liu, Fahri Ozturk, Shaopeng Huang, Qinquan Li, Huseyin Erdem Kazak, Gregoire Henry, Duo Wu
Summary: A lunar narrow field of view radiometer (LNR) has been designed by the Turkish Space Agency to measure short and long-wave radiation from a lunar orbiter. The LNR aims to achieve high-resolution and accurate lunar albedo and surface radiation measurement through optimized design.
2023 10TH INTERNATIONAL CONFERENCE ON RECENT ADVANCES IN AIR AND SPACE TECHNOLOGIES, RAST
(2023)
Article
Geochemistry & Geophysics
Huizeng Liu, Qingquan Li, Ping Zhu, Zhongwen Hu, Chao Yang, Yongquan Wang, Aihong Cui, Zuomin Wang, Guofeng Wu
Summary: This study investigates the potential application of Moon-based Earth observation (MEO) for monitoring the marine environment, particularly focusing on ocean color remote sensing. The results show that MEO-based remote sensing offers high spatial and temporal coverage, but there are limitations in high-latitude regions. Atmospheric and surface reflections affect sensor measurements, and further research is needed to improve accuracy. Overall, MEO-based ocean color remote sensing demonstrates great potential as a new perspective and long-term data source for marine environment monitoring.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Geochemistry & Geophysics
Huizeng Liu, Xianqiang He, Qingquan Li, Xianjun Hu, Joji Ishizaka, Susanne Kratzer, Chao Yang, Tiezhu Shi, Shuibo Hu, Qiming Zhou, Guofeng Wu
Summary: This study evaluated seven different atmospheric correction (AC) methods for the Ocean and Land Color Instrument (OLCI) on Sentinel-3. The results showed that different methods performed differently for different wavelength bands. POLYMER and C2RCC methods underestimated remote sensing reflectance at red and green bands, while SeaDAS method had an advantage for clear waters. Overall, POLYMER method performed best for chlorophyll retrieval.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)