The Erhai Lake Basin is an area with the active economic and social development of agriculture tourism, facing increasingly prominent environmental problems rapid urbanization. Assessing spatial–temporal changes in ecological environment quality objectively quantitatively a timely fashion crucial for protection policymaking. First, we selected high-quality Landsat imagery acquired at same time phase years 1999, 2004, 2009, 2014, 2019 respectively. Second, remote sensing-based index (RSEI)...

Timely and objective assessment of the optimal season for construction remote sensing ecological index (RSEI) is great significance accurate effective environment quality. We manipulated RSEI in to monitor seasonal variations quality (EEQ) Beijing-Tianjin-Hebei (JJJ) region from 2001 2020. First, we evaluated image across all four seasons filled missing observations through liner interpolation. Second, Seasonal was constructed using MODIS compared different years. Third, temporal spatial...

Vegetation productivity is crucial for human production and livelihoods. Understanding net primary (NPP) in historical contexts predicting its future fluctuations imperative assessing the environmental sustainability of a region. However, relatively few researches have been conducted on NPP, requiring further development refinement NPP prediction methods models. This study introduces novel approach that discretely couples PLUS CASA models prediction, it validates applicability this research...

<title>Abstract</title> African tropical forests have undergone extensive fragmentation, with an increasing proportion of previously intact now influenced by edge effects. It has become a pressing necessity to develop comprehensible index assess forest fragmentation and its interplay climate factors influencing ecosystem productivity (FEP). Using high-resolution cover maps, we developed Forest Fragmentation Gradient Index (FFGI), novel metric derived from two-dimensional framework...

Preseason temperature has always been considered the most critical factor influencing vegetation phenology in northern hemisphere. While numerous studies have examined impact of daytime and nighttime warming on this region, specific influence day night deciduous broad-leaved forests (DBFs) Northern China, where significant variations occur between night, remains unclear. Furthermore, sensitivity during different preseason periods to not quantitatively understood. We analyzed GIMMS3g NDVI...

The COVID-19 lockdown led to reduced industrial and transportation emissions in Chinese cities, improving air quality affecting large-scale vegetation. This study examines changes net primary productivity (NPP) across 283 prefecture-level cities China (PCC) during the lockdown, focusing on aerosol optical depth (AOD), nighttime light (NTL), temperature, precipitation. Results from spring 2020 show that 53.5% of experienced increased NPP, with greater gains high traffic activity due AOD....

The specific impact of ecological environment quality at a regional scale due to the rubber plantations expansion is still unclear in Xishuangbanna, Yunnan province, China. First, we used pixel and phenology-based multiple normalization approach map over six time periods during 1995–2018 with available high-quality Landsat imagery. Second, pixel-based optimized algorithm was developed reduce commission omission errors maps mapped area. Third, remote sensing-based index (RSEI) employed assess...

Tea plantations expansion has occurred in the major tea planting area of Yunnan province, China, however, it is a big challenge to extract and map distribution due their non-obvious phenological characteristics tropical subtropical regions. Firstly, we demonstrated pruning phase by using field hyperspectral data. Secondly, analyzed optical features typical land cover types from available time series Landsat imagery with Google Earth Engine (GEE) platform. We proposed pixel phenology- based...

Tea plantations encroachment into forests has occurred in the major tea planting area of Yunnan province, China past decades. However, dynamics plantation and effect on forest landscape pattern is not clear. In this study, we proposed a method to evaluate dynamic effects change by using time series Landsat 5/7/8 data from 1991 2021 Google Earth Engine (GEE) platform. First, pixel- phenology-based algorithm was applied generate maps for seven historical periods. Second, three indices...

The study of vegetation phenology and net primary production (NPP) plays a crucial role in understanding how ecosystems react to climate change. This monitored NPP Northeast China (NEC) from 2001 2022. results showed that: (1) Variations latitude longitude had highly significant effect (p<0.01) on all parameters. Besides, SOS (start growing season), EOS (end LOS (length season) NEC exhibited advancing (-0.26 days/a), delaying (0.14 prolonging (0.40 days/a) tendencies, respectively. Forest...

In order to improve the overall accuracy (OA) of AlexNet model for high-resolution remote sensing scene images with complex backgrounds, we proposed an improved image classification model. Firstly, used Layer Normalization (LN) replace Local Response (LRN) in and changed convolution kernel first layer 7 × 7. Secondly, focus on critical information feature extraction process, suppress irrelevant background information, two attention modules Convolution Block Attention Module (CBAM) Squeeze...

Timely and objective assessment of the optimal season for construction remote sensing ecological index (RSEI) is great significance accurate effective environment quality. We manipulated RSEI in different seasons to monitor evaluate seasonal quality (EEQ) variations Beijing-Tianjin-Hebei （JJJ） region from 2001 2020. First, we evaluated image quality, bad observations were interpolated. Second, was constructed by MODIS, compared eigenvalues contribution rate PC1 seasons. Third, assessed...

<title>Abstract</title> Context Vegetation productivity is crucial for human production and livelihoods. Monitoring changes in NPP (Net Primary Productivity) essential to evaluate regional ecological shifts carbon sink capacity. Objectives Our objective explore the variations of during 2001–2020 propose a new idea predict actual 2030 under multiple climate scenarios, taking Beijing-Tianjin-Hebei (BTH) region as an example. Methods This study utilized PLUS (patch-generating land use...

To improve model convergence speed and accuracy of the AlexNet on high-resolution remote sensing scene image classification, Batch Normalization (BN) layer were used to replace Local Response (LRN) normalize features each channel in convolution layers. In addition, we replaced filling method all layers with "SAME" reduce loss edge information convolution. Moreover, add dropout strategy after pooling prevent overfitting. Finally, three datasets including NWPU-RESISC45, AID, UCM for...

The tree canopy represents a fundamental element of tree-related information. However, achieving precise information from remote sensing images remains significant challenge due to varying sizes, mutual overlap, and diverse woodland environments. This study aims leverage high-resolution Chinese fir captured by an unmanned aerial vehicle (UAV) state forest farm in Jiangle County, Fujian Province, China. These are integrated with the Mask R-CNN model, automatically extract individual...

In order to prevent the spread of COVID-19, China implemented large-scale lockdown measures in early 2020. While studies have explored environmental changes during lockdowns, there has been a lack in-depth analysis regarding net primary productivity (NPP), and responses different vegetation types. The correlation between factors influenced by lockdowns remained unclear. this study, we evaluated spatial-temporal spring NPP at multiple scales period (LD, 2020) compared with unlocked (UL,...

A two-dimensional surrogate model technique of 3-D objects with varying material and shape is pretended. The based on gaussian process regression (GPR) constructs models to predict the monostatic RCS target. This method takes relative permittivity geometric size object as training inputs, while utilizes values outputs. series datasets are constructed using principles sampling. Through a test inputs involving variations in object, prediction results obtained from demonstrate high accuracy...