Self-Organizing Map for Erroneous Data Processing in Time Series Analysis

Identifying the primary factors influencing changes in vegetation cover within different land use policy zones, along with their spatial heterogeneity and the spatiotemporal evolution characteristics of past and future vegetation cover, is essential for effective Fractional Vegetation Cover (FVC) management. This study utilizes a Self-Organizing Feature Map (SOFM) neural network to identify land use policy zones in Anhui Province. By coupling the SEN+MK trend test with the Hurst index, the study predicts future vegetation change trends. A multiscale geographically weighted regression model is employed to reveal the spatial heterogeneity of the influencing factors of FVC, and a geographical detector is used to analyze the main influencing factors in different land use policy zones. The results indicate the following: From 2000 to 2023, the FVC in Anhui Province predominantly ranged between 0.4 and 0.6. Over these 24 years, the highest average FVC was found in the core ecological protection zone. The areas with the most significant improvement and fluctuation were located in the core agricultural production zone. Future improvements and degradations are primarily concentrated in the core agricultural production zone. The main driving factors differ across various land use policy zones. Climate factors dominate in Anhui Province as a whole, the core agricultural production zone, and the ecological protection transition zone. Topographical factors are predominant in the core ecological protection zone, whereas socioeconomic factors are the primary drivers in the core economic development zone. This study provides theoretical guidance for optimizing vegetation protection policies, coordinating human-environment relationships, and achieving sustainable development across different policy zones.

Abstract. Alpine meadow is one of the most widespread grassland types in the permafrost regions of the Qinghai-Tibet Plateau, and the transmission of coupled soil water heat is one of the most crucial processes influencing cyclic variations in the hydrology of frozen soil regions, especially under different vegetation covers. The present study assesses the impact of changes in vegetation cover on the coupling of soil water and heat in a permafrost region. Soil moisture (θv), soil temperature (Ts), soil heat content, and differences in θv−Ts coupling were monitored on a seasonal and daily basis under three different vegetation covers (30, 65, and 93%) on both thawed and frozen soils. Regression analysis of θv vs. Ts plots under different levels of vegetation cover indicates that soil freeze-thaw processes were significantly affected by the changes in vegetation cover. The decrease in vegetation cover of an alpine meadow reduced the difference between air temperature and ground temperature (ΔTa−s), and it also resulted in a decrease in Ts at which soil froze, and an increase in the temperature at which it thawed. This was reflected in a greater response of soil temperature to changes in air temperature (Ta). For ΔTa−s outside the range of −0.1 to 1.0°C, root zone soil-water temperatures showed a significant increase with increasing ΔTa−s; however, the magnitude of this relationship was dampened with increasing vegetation cover. At the time of maximum water content in the thawing season, the soil temperature decreased with increasing vegetation. Changes in vegetation cover also led to variations in θv−Ts coupling. With the increase in vegetation cover, the surface heat flux decreased. Soil heat storage at 20 cm in depth increased with increasing vegetation cover, and the heat flux that was downwardly transmitted decreased. The soil property varied greatly under different vegetation covers, causing the variation of heat conductivity and water-heat hold capacity in topsoil layer in different vegetation cover. The variation of heat budget and transmitting in soil is the main factor that causes changes in soil thawing and freezing processes, and θv−Ts coupling relationship under different vegetation fractions. In addition to providing insulation against soil warming, vegetation in alpine meadows within the permafrost region also would slow down the response of permafrost to climatic warming via the greater water-holding capacity of its root zone. Such vegetation may therefore play an important role in conserving water in alpine meadows and maintaining the stability of engineering works constructed within frozen soil of the Qinghai-Tibet Plateau.

Various neural network models have proven useful in evaluation or prediction. Neural classification ability is just beginning to be deployed in financial application. And it is very important to study credit evaluation model when create a credit risk prediction system. This paper analyses the disadvantage of traditional model based on statistical analysis, and proposed a hybrid system to combine the backpropagation (BP) learning with Kohonen's Self -Organizing Map (SOM) Neural Network, for the application of credit risk evaluation. BP Neural Network has been successfully used in several domains of artificial intelligence. In order to enhance its generalization performance, we connected the SOM method to deal with overfitting problem of BP. After discussing the structure and arithmetic of the model, we train the model with financial ratios for a credit risk early warning experiment. The preliminary experimental results demonstrate that the BP-SOM model outperforms some traditional ones in rates of prediction precision and efficiency, and improves generalization performance.

Segmentation of gray level images into regions of uniform texture is investigated. An unsupervised approach through the use of Kohonen's self organizing map (SOM) and a multilayer version of it, the hierarchical self organizing map (HSOM), is employed to find the regions in an image composed of textures from different classes. For testing, gray level artificial textured images modeled as Markov random fields are used as the input. No parameter estimation is done. The size and the topology of SOM and HSOM are independent from the size of the input image. The segmentation results are very promising.

In this paper, we investigate the features of spatiotemporal change in fractional vegetation cover (FVC) throughout the Yellow River Basin between 2000 and 2022 and identify the driving factors behind the change using the MODIS normalized difference vegetation index (NDVI) as a data source. On that basis, our research involves trend analysis and the center migration model to examine the correlation between vegetation cover changes and various factors, such as climates, topographies, soils, and human activities. In this way, we aim to uncover how such correlation contributed to the reduction in suspended sediment concentration (SSC) in surface runoff. Results suggest that (1) the FVC of the Yellow River Basin has been trending upward over the past 23 years, and vegetation growth has been remarkably improved; (2) the center of medium-high and denser vegetation cover has progressively migrated from the upper reaches of the Yellow River down to its middle-lower segments; (3) soil erosion control measures are critical to improve vegetation cover, given the great impact of the shifting natural elements on vegetation cover changes on a local basis; (4) an improvement in vegetation cover leads to considerable positive change in the runoff and SSC of the Yellow River mainstem. This study has yielded noteworthy contributions in correctly understanding the current vegetation changes and related factors in the Yellow River Basin. Furthermore, it has laid the groundwork for further research in improving the accuracy of basic data and deepening the correlation between factors.

Providing a powerful synchronization system is one of the most important goals to be pursued if an efficient utilization of sensor networks has to be addressed. The basic concept behind a Wireless Sensor Network is to deploy a large number of sensor nodes able to acquire and process data. Most of WSNs applications require sensor nodes to maintain local clocks both to determine the events order and to provide temporal information to measured data; to achieve this goal applications generally require a low synchronization precision, close to Milli seconds. This paper proposes a novel synchronization system based on Kohonen's Self Organizing Maps (SOMs), able to provide some Artificial Intelligence features to sensor nodes. A SOM is a particular neural network that learns to classify data without any supervision. In each sensor node, a SOM is implemented to evaluate the sensor node time, using a very little amount of storage and computing resources. In a scenario where thousands of sensor nodes are placed, this system evaluates the time of each sensor in a distributed manner, assuming a very little percentage of nodes knowing the actual time, thus ensuring an effective clock synchronization among all the sensors.

Spatiotemporal variations of Chinese Loess Plateau vegetation cover during 1981-2006 have been investigated using GIMMS and SPOT VGT NDVI data and the cause of vegetation cover changes has been analyzed, considering the climate changes and human activities. Vegetation cover changes on the Loess Plateau have experienced four stages as follows: (1) vegetation cover showed a continued increasing phase during 1981-1989; (2) vegetation cover changes came into a relative steady phase with small fluctuations during 1990-1998; (3) vegetation cover declined rapidly during 1999-2001; and (4) vegetation cover increased rapidly during 2002-2006. The vegetation cover changes of the Loess Plateau show a notable spatial difference. The vegetation cover has obviously increased in the Inner Mongolia and Ningxia plain along the Yellow River and the ecological rehabilitated region of Ordos Plateau, however the vegetation cover evidently decreased in the hilly and gully areas of Loess Plateau, Liupan Mountains region and the northern hillside of Qinling Mountains. The response of NDVI to climate changes varied with different vegetation types. NDVI of sandy land vegetation, grassland and cultivated land show a significant increasing trend, but forest shows a decreasing trend. The results obtained in this study show that the spatiotemporal variations of vegetation cover are the outcome of climate changes and human activities. Temperature is a control factor of the seasonal change of vegetation growth. The increased temperature makes soil drier and unfavors vegetation growth in summer, but it favors vegetation growth in spring and autumn because of a longer growing period. There is a significant correlation between vegetation cover and precipitation and thus, the change in precipitation is an important factor for vegetation variation. The improved agricultural production has resulted in an increase of NDVI in the farmland, and the implementation of large-scale vegetation construction has led to some beneficial effect in ecology.

Monitoring changes in vegetation cover is important for the restoration of ecosystems in the Gorontalo Regency area. The utilization of remote sensing technology makes it possible to detect the dynamics of changes in vegetation cover spatially and temporally. The Terra MODIS satellite image collection in the study area is available in large numbers and sizes. Therefore, cloud computing-based spatial technology support is needed. Google Earth Engine (GEE) as a geospatial computing device is an alternative to cover this shortfall. The aim of this study is to explore the condition of vegetation cover spatially and temporally using the GEE platform. A total of 43 MODIS images in the study area, recording periods 2000 and 2020, were used to quickly and effectively generate vegetation cover maps. The process of downloading, processing, and analyzing data was automated through the GEE interface. The results of the mapping in 2000 and 2020 are shown by maps of vegetation cover in two classes, namely; vegetation and non-vegetation. The accuracy of the vegetation cover map shows good results, namely an overall accuracy of 0.81 for 2000 and 0.85 for 2020. The area of the non-vegetation class increased by 2815.29 ha, and the vegetation class decreased by 2767.31 ha. The map of spatial changes in vegetation cover in the study area is classified into three classes, namely revegetation, devegetation, and unchanged. Based on these results, the extraction of vegetation cover changes in the study area using the GEE platform can be carried out well.

The earth is the home of human survival, and vegetation, as an important natural resource on the land surface, has the functions of sand sealing and soil consolidation, water conservation and so on. Change of vegetation can reflect the change of ecological environment. It is also an important role in the energy cycle and water cycle of the earth's surface and atmosphere. By studying the change of vegetation coverage, we can get the information of ecology and water source. Then, we need to take effective measures to improve our ecological environment. It is also of great significance for us to understand the ecological environment. In this paper, the clear images of 1990, 1998, 2015and 2018in June and July are downloaded from Landsat 8 satellite image data. The vegetation index NDVI of each year is calculated by ArcMap software, and the vegetation coverage is calculated by NDVI. The changes of NDVI and vegetation coverage are analysed. The influencing factors of vegetation coverage and the objective control methods are put forward. Based on the change of vegetation coverage from 1990to 2018, it is concluded that the area of vegetation restoration in the upper reaches of Shiyang River Basin is larger than that of degradation, that in the middle reaches of the basin is larger than that of degradation that in the lower reaches of the basin is larger than that of vegetation restoration, and that in Minqin Basin is larger than that of vegetation restoration Product. Generally speaking, the area of vegetation restoration in Shiyang River Basin is slightly larger than that of vegetation degradation. Based on the statistical analysis of the vegetation coverage in the basin, we should focus on the control of the vegetation in the extremely low vegetation coverage area and the medium vegetation coverage area.

Changes in vegetation cover are known to influence the climate system by modifying the radiative, momentum, and hydrologic balance of the land surface. To explore the interactions between terrestrial vegetation and the atmosphere for doubled atmospheric CO2 concentrations, the newly developed fully coupled GENESIS–IBIS climate–vegetation model is used. The simulated climatic response to the radiative and physiological effects of elevated CO2 concentrations, as well as to ensuing simulated shifts in global vegetation patterns is investigated. The radiative effects of elevated CO2 concentrations raise temperatures and intensify the hydrologic cycle on the global scale. In response, soil moisture increases in the mid- and high latitudes by 4% and 5%, respectively. Tropical soil moisture, however, decreases by 5% due to a decrease in precipitation minus evapotranspiration. The direct, physiological response of plants to elevated CO2 generally acts to weaken the earth’s hydrologic cycle by lowering transpiration rates across the globe. Lowering transpiration alone would tend to enhance soil moisture. However, reduced recirculation of water in the atmosphere, which lowers precipitation, leads to more arid conditions overall (simulated global soil moisture decreases by 1%), particularly in the Tropics and midlatitudes. Allowing structural changes in the vegetation cover (in response to changes in climate and CO2 concentrations) overrides the direct physiological effects of CO2 on vegetation in many regions. For example, increased simulated forest cover in the Tropics enhances canopy evapotranspiration overall, offsetting the decreased transpiration due to lower leaf conductance. As a result of increased circulation of moisture through the hydrologic cycle, precipitation increases and soil moisture returns to the value simulated with just the radiative effects of elevated CO2. However, in the highly continental midlatitudes, changes in vegetation cover cause soil moisture to decline by an additional 2%. Here, precipitation does not respond sufficiently to increased plant-water uptake, due to a limited source of external moisture into the region. These results illustrate that vegetation feedbacks may operate differently according to regional characteristics of the climate and vegetation cover. In particular, it is found that CO2 fertilization can cause either an increase or a decrease in available soil moisture, depending on the associated changes in vegetation cover and the ability of the regional climate to recirculate water vapor. This is in direct contrast to the view that CO2 fertilization will enhance soil moisture and runoff across the globe: a view that neglects changes in vegetation structure and local climatic feedbacks.

It is generally believed that there is an inverse relationship between population growth and vegetation cover. However, reports about vegetation protection and reforestation around the World have been continuously increasing in recent decades, which seems to indicate that this relationship may not be true. In this paper, we have taken 21 cities in Guangdong Province, China as the study area to test the long-term relationship between population growth and vegetation cover, using an AVHRR NDVI data set and the panel cointegrated regression method. The results show that there is a long-term inverted N-shaped curve relationship between population growth and vegetation cover in the region where there are frequent human activities and the influence of climate change on vegetation cover changes is relatively small. The two turning points of the inverted N-shaped curve for the case of Guangdong Province correspond to 2,200 persons·km−2 and 3,820 persons·km−2, and they can provide a reference range for similar regions of the World. It also states that the population urbanization may have a negative impact on the vegetation cover at the early stage, but have a positive impact at the later stage. In addition, the Panel Error Correction Model (PECM) is used to investigate the causality direction between population growth and vegetation cover. The results show that not only will the consuming destruction effect and planting construction effect induced by the population growth have a great impact on vegetation cover changes, but vegetation cover changes in turn will also affect the population growth in the long term.

Offshore wind turbine (OWT) installations are continually expanding as they are considered an efficient mechanism for covering a part of the energy consumption requirements. The assessment of the energy potential of OWTs for specific offshore sites is the key factor that defines their successful implementation, commercialization and sustainability. The data used for this assessment mainly refer to wind speed measurements. However, the data may not present homogeneity due to incomplete or missing entries; this in turn, is attributed to failures of the measuring devices or other factors. This fact may lead to considerable limitations in the OWTs energy potential assessment. This paper presents two novel methodologies to handle the problem of incomplete and missing data. Computational intelligence algorithms are utilized for the filling of the incomplete and missing data in order to build complete wind speed series. Finally, the complete wind speed series are used for assessing the energy potential of an OWT in a specific offshore site. In many real-world metering systems, due to meter failures, incomplete and missing data are frequently observed, leading to the need for robust data handling. The novelty of the paper can be summarized in the following points: (i) a comparison of clustering algorithms for extracting typical wind speed curves is presented for the OWT related literature and (ii) two efficient novel methods for missing and incomplete data are proposed.

This study was conducted to nondestructively estimate the volumetric information of peach and pit and to visualize the 3D information of internal structure from magnetic resonance imaging(MRI) data. Bruker Biospec 7T spectrometer operating at a proton reosonant frequency of 300 MHz was used for acquisition of MRI data of peach. Image processing algorithms and visualization techniques were implemented by using MATLAB (Mathworks) and Visualization Toolkit(Kitware), respectively. Thresholding algorithm and Kohonen's self organizing map(SOM) were applied to MRI data fur region segmentation. Volumetric information were estimated from segemented images and compared to the actual measurements. The average prediction errors of peach and pit volumes were 4.5%, 26.1%, respectively for the thresholding algorithm. and were 2.1%, 19.9%. respectively for the SOM. Although we couldn't get the statistically meaningful results with the limited number of samples, the average prediction errors were lower when the region segmentation was done by SOM rather than thresholding. The 3D visualization techniques such as isosurface construction and volume rendering were successfully implemented, by which we could nondestructively obtain the useful information of internal structures of peach.

Since their introduction by Kohonen Self Organizing Maps (SOMs) have been used in various forms for purposes of surface reconstruction. They offer robust and fast approximations of manifold data from unstructured input points while being modestly easy to implement. On the other hand SOMs have certain disadvantages when used in a setup where sparse, reliable and spacial unbounded data occurs. For example, airborne Lidar sensors generate a continuous stream of point data while flying above terrain. We introduce modifications of the SOM's data structure to adapt it to unbounded data. Furthermore, we introduce a new variation of the learning rule called rapid learning that is feasible for sparse but rather reliable data. We demonstrate examples where the surroundings of an aircraft can be reconstructed in almost real time.

The availability of suitable breeding habitats is an important factor in explaining seabird demographic fluctuations. For ground nesting seabirds, such as Little Terns (Sternula albifrons), an essential characteristic in determining suitability of breeding habitats is vegetative cover. This study evaluated whether the long-term fluctuations (1979–2014) in Little Tern breeding numbers on sandy beach colonies of Ria Formosa, Algarve, Portugal, could be explained by changes in vegetation cover at each sandy beach colony, environmental variability (measured by the climatic North Atlantic Oscillation index), or a combination of both. The percentage of vegetation cover for each sandy beach was calculated from aerial photos of Ria Formosa barrier islands from 1976 to 2012, and related to breeding census data from the same period. The climatic variation measured by the North Atlantic Oscillation index did not show a significant relationship with the number of breeding pairs nesting on each sandy beach colony. A segmented regression model constructed to model the effect of vegetation cover on the number of breeding pairs on sandy beaches found an optimal vegetation cover of around 10% and an overall breakpoint of 42%vegetation cover; the model predicted that only ∼4% of breeding pairs would endure with 90% vegetation cover and none would endure with 92% vegetation cover. Little Terns in Ria Formosa appear to adopt a fugitive strategy, moving their colonies to locations with better breeding conditions in response to the overgrowth of vegetation.