Zipeng Zhao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Zipeng Zhao has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Fan Wang
Department of Forest Management, School of Forestry, Northeast Forestry University, Harbin 150040, China
3 shared publications - Weiwei Jia
Department of Forest Management, School of Forestry, Northeast Forestry University, Harbin, China
3 shared publications - Yuman Sun
Department of Forest Management, School of Forestry, Northeast Forestry University, Harbin 150040, China
3 shared publications - Haiping Zhao
Department of Forest Management, School of Forestry, Northeast Forestry University, Harbin 150040, China
2 shared publications - Haotian Guo
Department of Forest Management, School of Forestry, Northeast Forestry University, Harbin 150040, China
1 shared publication - Jin-Yan Yang
College of Architecture and Environment, Sichuan University, Chengdu, People’s Republic of China
1 shared publication
Latest Publications
- Open Access
- 25 March 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17 (7) , 1164
Prediction of Vanadium Contamination Distribution Pattern Through Remote Sensing Image Fusion and Machine Learning
Soil vanadium contamination poses a significant threat to ecosystems. Hyperspectral remote sensing plays a critical role in extracting spectral features of heavy metal contamination, mapping its spatial distribution, and monitoring its trends over time. This study targets a vanadium-contaminated area in Panzhihua City, Sichuan Province. Soil sampling and spectral measurements occurred in the laboratory. Hyperspectral (Gaofen-5, GF-5) and multispectral (Gaofen-2, GF-2; Sentinel-2) images were acquired and preprocessed, and feature bands were extracted by combining laboratory spectral data. A dual-branch convolutional neural network (DB-CNN) fused hyperspectral and multispectral images and confirmed the fusion’s effectiveness. Six prevalent machine learning models were adopted, and a unified learning framework leveraged a Random Forest (RF) as a second-layer model to enhance the predictive performance of these base models. Both the base models and the ensemble learning model were evaluated based on predictive accuracy. The fusion process enhanced the predictive performance of the base models, improving R2 values for vanadium (V) and pentavalent vanadium (V5+) from 0.54 and 0.3 to 0.58 and 0.39, respectively, at a 4 m resolution. Further optimization using RF as a second-layer model to refine Extreme Trees (ETs) significantly increased R2 values to 0.83 and 0.75 for V and V5+, respectively, at this scale. The 934 nm and 464 nm wavelengths were identified as the most critical spectral bands for predicting soil vanadium contamination. This integrated approach robustly delineates the spatial distribution characteristics of V and V5+ in soils, facilitating precise monitoring and ecological risk assessments of vanadium contamination through a comparative analysis of predictive accuracy across diverse models.
- 25 March 2025
- Published by MDPI AG in Remote Sensing
- Vol. 17 (7) , 1164
Open AccessPrediction of Vanadium Contamination Distribution Pattern Through Remote Sensing Image Fusion and Machine Learning
Soil vanadium contamination poses a significant threat to ecosystems. Hyperspectral remote sensing plays a critical role in extracting spectral features of heavy metal contamination, mapping its spatial distribution, and monitoring its trends over time. This study targets a vanadium-contaminated area in Panzhihua City, Sichuan Province. Soil sampling and spectral measurements occurred in the laboratory. Hyperspectral (Gaofen-5, GF-5) and multispectral (Gaofen-2, GF-2; Sentinel-2) images were acquired and preprocessed, and feature bands were extracted by combining laboratory spectral data. A dual-branch convolutional neural network (DB-CNN) fused hyperspectral and multispectral images and confirmed the fusion’s effectiveness. Six prevalent machine learning models were adopted, and a unified learning framework leveraged a Random Forest (RF) as a second-layer model to enhance the predictive performance of these base models. Both the base models and the ensemble learning model were evaluated based on predictive accuracy. The fusion process enhanced the predictive performance of the base models, improving R2 values for vanadium (V) and pentavalent vanadium (V5+) from 0.54 and 0.3 to 0.58 and 0.39, respectively, at a 4 m resolution. Further optimization using RF as a second-layer model to refine Extreme Trees (ETs) significantly increased R2 values to 0.83 and 0.75 for V and V5+, respectively, at this scale. The 934 nm and 464 nm wavelengths were identified as the most critical spectral bands for predicting soil vanadium contamination. This integrated approach robustly delineates the spatial distribution characteristics of V and V5+ in soils, facilitating precise monitoring and ecological risk assessments of vanadium contamination through a comparative analysis of predictive accuracy across diverse models.
- Open Access
- 8 October 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (19) , 4869
Study on the Regeneration Probability of Understory Coniferous Saplings in the Liangshui Nature Reserve Based on Four Modeling Techniques
Forests are one of the most important natural resources for humans, and understanding the regeneration probability of undergrowth in forests is very important for future forest spatial structure and forest management. In addition, the regeneration of understory saplings is a key process in the restoration of forest ecosystems. By studying the probability of sapling regeneration in forests, we can understand the impact of different stand factors and environmental factors on sapling regeneration. This could help provide a scientific basis for the restoration and protection of forest ecosystems. The Liangshui Nature Reserve of Yichun City, Heilongjiang Province, is a coniferous and broadleaved mixed forest. In this study, we assess the regeneration probability of coniferous saplings (CRP) in natural forests in 665 temporary plots in the Liangshui Nature Reserve. Using Sentinel-1 and Sentinel-2 images provided by the European Space Agency, as well as digital elevation model (DEM) data, we calculated the vegetation index, microwave vegetation index (RVI S1), VV, VH, texture features, slope, and DEM and combined them with field survey data to construct a logistic regression (LR) model, geographically weighted logistic regression (GWLR) model, random forest (RF) model, and multilayer perceptron (MLP) model to predict and analyze the CRP value of each pixel in the study area. The accuracy of the models was evaluated with the average values of the area under the ROC curve (AUC), kappa coefficient (KAPPA), root mean square error (RMSE), and mean absolute error (MAE) verified by five-fold cross-validation. The results showed that the RF model had the highest accuracy. The variable factor with the greatest impact on CRP was the DEM. The construction of the GWLR model considered more spatial factors and had a lower residual Moran index value. The four models had higher CRP prediction results in the low-latitude and low-longitude regions of the study area, and in the high-latitude and high-longitude regions of the study area, most pixels had a CRP value of 0 (i.e., no coniferous sapling regeneration occurred).
- 8 October 2023
- Published by MDPI AG in Remote Sensing
- Vol. 15 (19) , 4869
Open AccessStudy on the Regeneration Probability of Understory Coniferous Saplings in the Liangshui Nature Reserve Based on Four Modeling Techniques
Forests are one of the most important natural resources for humans, and understanding the regeneration probability of undergrowth in forests is very important for future forest spatial structure and forest management. In addition, the regeneration of understory saplings is a key process in the restoration of forest ecosystems. By studying the probability of sapling regeneration in forests, we can understand the impact of different stand factors and environmental factors on sapling regeneration. This could help provide a scientific basis for the restoration and protection of forest ecosystems. The Liangshui Nature Reserve of Yichun City, Heilongjiang Province, is a coniferous and broadleaved mixed forest. In this study, we assess the regeneration probability of coniferous saplings (CRP) in natural forests in 665 temporary plots in the Liangshui Nature Reserve. Using Sentinel-1 and Sentinel-2 images provided by the European Space Agency, as well as digital elevation model (DEM) data, we calculated the vegetation index, microwave vegetation index (RVI S1), VV, VH, texture features, slope, and DEM and combined them with field survey data to construct a logistic regression (LR) model, geographically weighted logistic regression (GWLR) model, random forest (RF) model, and multilayer perceptron (MLP) model to predict and analyze the CRP value of each pixel in the study area. The accuracy of the models was evaluated with the average values of the area under the ROC curve (AUC), kappa coefficient (KAPPA), root mean square error (RMSE), and mean absolute error (MAE) verified by five-fold cross-validation. The results showed that the RF model had the highest accuracy. The variable factor with the greatest impact on CRP was the DEM. The construction of the GWLR model considered more spatial factors and had a lower residual Moran index value. The four models had higher CRP prediction results in the low-latitude and low-longitude regions of the study area, and in the high-latitude and high-longitude regions of the study area, most pixels had a CRP value of 0 (i.e., no coniferous sapling regeneration occurred).
- Open Access
Comparison of Global and Local Poisson Models for the Number of Recruitment Trees in Natural Forests
The recruitment of natural forests is the key to stand growth and regeneration. Constructing theoretical models for recruitment trees is crucial for accurately quantifying stand growth and yield. To this end, the objective was to use relevant Poisson models to study the spatial relationships between the number of recruitment trees (NRTs) and driving factors, such as topography, stand, and remote sensing factors. Taking the Northeast China Liangshui Nature Reserve as the study area and 127 ecological public welfare forest plots based on grid sampling as study data, we constructed global models (Poisson regression (PR) and linear mixed Poisson regression (LMPR)) and local models (geographically weighted Poisson regression (GWPR) and semiparametric GWPR (SGWPR)) to simulate the NRTs. The evaluation indicators were calculated to analyse four model fittings, predictive abilities, and spatial effects of residual analysis. The results show that local (GWPR and SGWPR) models have great advantages in all aspects. Compared with the GWPR model, the SGWPR model exhibited improved performance by considering whether coefficients have geographical variability for all independent variables. Therefore, the SGWPR model more accurately depicts the spatial distributions of NRTs than the other models.
Open AccessComparison of Global and Local Poisson Models for the Number of Recruitment Trees in Natural Forests
The recruitment of natural forests is the key to stand growth and regeneration. Constructing theoretical models for recruitment trees is crucial for accurately quantifying stand growth and yield. To this end, the objective was to use relevant Poisson models to study the spatial relationships between the number of recruitment trees (NRTs) and driving factors, such as topography, stand, and remote sensing factors. Taking the Northeast China Liangshui Nature Reserve as the study area and 127 ecological public welfare forest plots based on grid sampling as study data, we constructed global models (Poisson regression (PR) and linear mixed Poisson regression (LMPR)) and local models (geographically weighted Poisson regression (GWPR) and semiparametric GWPR (SGWPR)) to simulate the NRTs. The evaluation indicators were calculated to analyse four model fittings, predictive abilities, and spatial effects of residual analysis. The results show that local (GWPR and SGWPR) models have great advantages in all aspects. Compared with the GWPR model, the SGWPR model exhibited improved performance by considering whether coefficients have geographical variability for all independent variables. Therefore, the SGWPR model more accurately depicts the spatial distributions of NRTs than the other models.