Identification of Bamboo Species Based on Extreme Gradient Boosting (XGBoost) Using Zhuhai-1 Orbita Hyperspectral Remote Sensing Imagery.

Mapping the distribution of bamboo species is vital for the sustainable management of bamboo and for assessing its ecological and socioeconomic value. However, the spectral similarity between bamboo species makes this work extremely challenging through remote sensing technology. Existing related studies rarely integrate multiple feature variables and consider how to quantify the main factors affecting classification. Therefore, feature variables, such as spectra, topography, texture, and vegetation indices, were used to construct the XGBoost model to identify bamboo species using the Zhuhai-1 Orbita hyperspectral (OHS) imagery in the Southern Sichuan Bamboo Sea and its surrounding areas in Sichuan Province, China. The random forest and Spearman's rank correlation analysis were used to sort the main variables that affect classification accuracy and minimize the effects of multicollinearity among variables. The main findings were: (1) The XGBoost model achieved accurate and reliable classification results. The XGBoost model had a higher overall accuracy (80.6%), kappa coefficient (0.708), and mean F1-score (0.805) than the spectral angle mapper (SAM) method; (2) The optimal feature variables that were important and uncorrelated for classification accuracy included the blue band (B1, 464-468 nm), near-infrared band (B27, 861-871 nm), green band (B5, 534-539 nm), elevation, texture feature mean, green band (B4, 517-523 nm), and red edge band (B17, 711-720 nm); and (3) the XGBoost model based on the optimal feature variable selection showed good adaptability to land classification and had better classification performance. Moreover, the mean F1-score indicated that the model could well balance the user's and producer's accuracy. Additionally, our study demonstrated that OHS imagery has great potential for land cover classification and that combining multiple features to enhance classification is an approach worth exploring. Our study provides a methodological reference for the application of OHS images for plant species identification.

Spatial-Temporal Evolution and Driving Forces of Drying Trends on the Qinghai-Tibet Plateau Based on Geomorphological Division.

The Qinghai-Tibet Plateau (QTP) is a sensor of global climate change and regional human activities, and drought monitoring will help to achieve its ecological protection and sustainable development. In order to effectively control the geospatial scale effect, we divided the study area into eight geomorphological sub-regions, and calculated the Temperature-Vegetation Drought Index (TVDI) of each geomorphological sub-region based on MODIS Normalized Difference Vegetation Index (NDVI) and Land Surface Temperature (LST) data, and synthesized the TVDI of the whole region. We employed partial and multiple correlation analyses to identify the relationship between TVDI and temperature and precipitation. The random forest model was further used to study the driving mechanism of TVDI in each geomorphological division. The results of the study were as follows: (1) From 2000 to 2019, the QTP showed a drought trend, with the most significant drought trend in the central region. The spatial pattern of TVDI changes of QTP was consistent with the gradient changes of precipitation and temperature, both showing a gradual trend from southeast to northwest. (2) There was a risk of drought in the four seasons of the QTP, and the seasonal variation of TVDI was significant, which was characterized by being relatively dry in spring and summer and relatively humid in autumn and winter. (3) Drought in the QTP was mainly driven by natural factors, supplemented by human factors. The driving effect of temperature and precipitation factors on TVDI was stable and significant, which mainly determined the spatial distribution and variation of TVDI of the QTP. Geomorphological factors led to regional intensification and local differentiation effects of drought, especially in high mountains, flat slopes, sunny slopes and other places, which had a more significant impact on TVDI. Human activities had local point-like and linear impacts, and grass-land and cultivated land that were closely related to the relatively high impacts on TVDI of human grazing and farming activities. In view of the spatial-temporal patterns of change in TVDI in the study area, it is important to strengthen the monitoring and early warning of changes in natural factors, optimize the spatial distribution of human activities, and scientifically promote ecological protection and restoration.

Spatial-Temporal Evolution and Driving Forces of NDVI in China's Giant Panda National Park.

Identifying the ecological evolution trends and vegetation driving mechanisms of giant panda national parks can help to improve the protection of giant panda habitats. Based on the research background of different geomorphological zoning, we selected the MODIS NDVI data from 2000 to 2020 to analyze the NDVI trends using a univariate linear model. A partial correlation analysis and multiple correlation analysis were used to reveal the influence of temperature and precipitation on NDVI trends. Fourteen factors related to meteorological factors, topographic factors, geological activities, and human activities were selected, and the Geographically Weighted Regression model was used to study the mechanisms driving NDVI change. The results were as follows: (1) The NDVI value of Giant Panda National Park has fluctuated and increased in the past 21 years, with an annual growth rate of 4.7%/yr. Affected by the Wenchuan earthquake in 2008, the NDVI value fluctuated greatly from 2008 to 2012, and reached its peak in 2018. (2) The NDVI in 94% of the study area improved, and the most significant improvement areas were mainly distributed in the northern and southern regions of Southwest Subalpine and Middle Mountain and the Xiaoxiangling area. Affected by the distribution of fault zones and their local activities, vegetation degradation was concentrated in the Dujiangyan-Anzhou area of Hengduan Mountain Alpine Canyon. (3) The Geographically Weighted Regression analysis showed that natural factors were dominant, with climate and elevation having a double-factor enhancement effect, the peak acceleration of ground motion and fault zone having a superimposed effect, and river density and slope having a double effect, all of which had a significant impact on the NDVI value of the surrounding area. To optimize the ecological security pattern of the Giant Panda National Park, we recommended strengthening the construction of ecological security projects through monitoring meteorological changes, preventing, and controlling geo-hazards, and optimizing the layout and intensity of human activities.

Susceptibility Analysis of Geohazards in the Longmen Mountain Region after the Wenchuan Earthquake.

Multitemporal geohazard susceptibility analysis can not only provide reliable results but can also help identify the differences in the mechanisms of different elements under different temporal and spatial backgrounds, so as to better accurately prevent and control geohazards. Here, we studied the 12 counties (cities) that were severely affected by the Wenchuan earthquake of 12 May 2008. Our study was divided into four time periods: 2008, 2009-2012, 2013, and 2014-2017. Common geohazards in the study area, such as landslides, collapses and debris flows, were taken into account. We constructed a geohazard susceptibility index evaluation system that included topography, geology, land cover, meteorology, hydrology, and human activities. Then we used a random forest model to study the changes in geohazard susceptibility during the Wenchuan earthquake, the following ten years, and its driving mechanisms. We had four main findings. (1) The susceptibility of geohazards from 2008 to 2017 gradually increased and their spatial distribution was significantly correlated with the main faults and rivers. (2) The Yingxiu-Beichuan Fault, the western section of the Jiangyou-Dujiangyan Fault, and the Minjiang and Fujiang rivers were highly susceptible to geohazards, and changes in geohazard susceptibility mainly occurred along the Pingwu-Qingchuan Fault, the eastern section of the Jiangyou-Dujiangyan Fault, and the riparian areas of the Mianyuan River, Zagunao River, Tongkou River, Baicao River, and other secondary rivers. (3) The relative contribution of topographic factors to geohazards in the four different periods was stable, geological factors slowly decreased, and meteorological and hydrological factors increased. In addition, the impact of land cover in 2008 was more significant than during other periods, and the impact of human activities had an upward trend from 2008 to 2017. (4) Elevation and slope had significant topographical effects, coupled with the geological environmental effects of engineering rock groups and faults, and river-derived effects, which resulted in a spatial aggregation of geohazard susceptibility. We attributed the dynamic changes in the areas that were highly susceptible to geohazards around the faults and rivers to the changes in the intensity of earthquakes and precipitation in different periods.

[Spectral Response and Inversion Models for Prediction of Total Copper Content in Soil of Xifanping Mining Area].

In order to solve the problem of high cost and low efficiency by using the traditional soil geochemical survey methods, this paper studied the simple detection method of soil heavy metal content with visible and near-infrared reflectance spectroscopy. The study collected visible and near-infrared reflectance spectroscopy of soil samples in Xifanping mining area; then treated the reflectance spectroscopy with six mathematic changes such as differentials and continuum removal in advance; the next step was to select characteristic wavelengths that were sensitive to soil copper content by using stepwise regression method and Pearson correlation coefficient as set of comprehensive characteristic variables; finally, utilized different methods and parameters of characteristic variable selection to build the soil total copper content models and tested them. Results showed that: to extract the information of copper content in soil, the performance of different spectral transform methods varied, and each spectrum transform method corresponded to its certain sensitive spectral ranges; the inversion models based on the integrated spectrum transform information were better than that based on only one kind of spectrum transform information; as for establishing the prediction model of soil copper content by using the integrated spectrum transform information, backward elimination was better than forward selection and stepwise selection, and when the Removal is 0.20, the optimum model was obtained, its coefficients of determination(R(2))and determination coefficients of prediction(R(2)(pre))reached 0.851 and 0.830, root mean square error of calibration(RMSEC)and root mean square error of prediction(RMSEP)were 0.349 and 0.468 mg·kg(-1). The model has a good precision, and it provides a train of thought for the detection of other soil heavy metal elements with visible and near-infrared reflectance spectroscopy.