Local Spatial Autocorrelation Analysis Research Articles

When Wurst comes to Wurscht: Variation and koiné formation in Texas German

AbstractTexas German is a new world language variety that shows some evidence of koiné development but also presents with substantial variation at many levels of structure. I present a case study on the variant pronunciation of sibilants in Texas German consonant clusters. This feature is fairly frequent and found throughout the regions of German settlement in Central Texas. After a discussion of the presence of this feature in the donor dialects, I investigate the factors that correlate with variation in the modern language. From an analysis of local and global spatial autocorrelation, I argue that variation is not significantly associated with particular geographic regions and is compatible with stable and homogenous variation. This provides insight into our understanding of new dialect emergence and the mechanisms by which dialect features are leveled over multiple generations.

Socioeconomic and environmental factors associated with malaria hotspots in the Nanoro demographic surveillance area, Burkina Faso

BackgroundWith limited resources and spatio-temporal heterogeneity of malaria in developing countries, it is still difficult to assess the real impact of socioeconomic and environmental factors in order to set up targeted campaigns against malaria at an accurate scale. Our goal was to detect malaria hotspots in rural area and assess the extent to which household socioeconomic status and meteorological recordings may explain the occurrence and evolution of these hotspots.MethodsData on malaria cases from 2010 to 2014 and on socioeconomic and meteorological factors were acquired from four health facilities within the Nanoro demographic surveillance area. Statistical cross correlation was used to quantify the temporal association between weekly malaria incidence and meteorological factors. Local spatial autocorrelation analysis was performed and restricted to each transmission period using Kulldorff’s elliptic spatial scan statistic. Univariate and multivariable analysis were used to assess the principal socioeconomic and meteorological determinants of malaria hotspots using a Generalized Estimating Equation (GEE) approach.ResultsRainfall and temperature were positively and significantly associated with malaria incidence, with a lag time of 9 and 14 weeks, respectively. Spatial analysis showed a spatial autocorrelation of malaria incidence and significant hotspots which was relatively stable throughout the study period. Furthermore, low socioeconomic status households were strongly associated with malaria hotspots (aOR = 1.21, 95% confidence interval: 1.03–1.40).ConclusionThese fine-scale findings highlight a relatively stable spatio-temporal pattern of malaria risk and indicate that social and environmental factors play an important role in malaria incidence. Integrating data on these factors into existing malaria struggle tools would help in the development of sustainable bottleneck strategies adapted to the local context for malaria control.

Examining the spatiotemporal evolution of vaccine refusal: nonmedical exemptions from vaccination in California, 2000\u20132013

BackgroundVaccine hesitancy continues to be an issue throughout the United States, as numerous vaccine hesitant parents are choosing to exempt their children from school-entry vaccination requirements for nonmedical reasons, despite the safety and effectiveness of vaccines. We conduct an analysis of how vaccine refusal, measured by the use of nonmedical exemptions (based on personal or religious beliefs) from vaccination (NMEs), evolved across space and over time in California.MethodsUsing school-entry data from the California Department of Public Health, we examined NMEs for students entering kindergarten in California from 2000 to 2013. We conduct global and local spatial autocorrelation analysis to determine whether NME use became more geographically clustered over the study period and whether the location of local clusters of high use were temporally stable. We conducted a grouping analysis that identified the general temporal trends in NME use over the time period.ResultsThe use of NMEs increased from 0.73% of all kindergarteners in 2000 to 3.09% in 2013 and became more geographically clustered over the study period. Local geographic clusters of high use were relatively isolated early in the study period, but expanded in size over time. The grouping analysis showed that regions with high NME use early in the study period were generally few (15% of all US Census tracts) and relatively isolated. Regions that had low initial NME use and moderate to large increases over the study period were located in close proximity to the initial high use regions. The grouping analysis also showed that roughly half of all tracts had 0% or very low NME use throughout the study period.ConclusionsWe found an observable spatial structure to vaccine refusal and NME use over time, which appeared to be a self-reinforcing process, as well as a spatially diffusive process. Importantly, we found evidence that use of NMEs in the initially isolated regions appeared to stimulate vaccine refusal in geographically proximal regions. Thus, our results suggest that efforts aimed at decreasing future NME use may be most effective if they target regions where NME use is already high, as well as the nearby regions.

Spatiotemporal patterns of severe fever with thrombocytopenia syndrome in China, 2011–2016

Severe fever with thrombocytopenia syndrome (SFTS) is emerging and the number of SFTS cases have increased year by year in China. However, spatiotemporal patterns and trends of SFTS are less clear up to date. In order to explore spatiotemporal patterns and predict SFTS incidences, we analyzed temporal trends of SFTS using autoregressive integrated moving average (ARIMA) model, spatial patterns, and spatiotemporal clusters of SFTS cases at the county level based on SFTS data in China during 2011–2016. We determined the optimal time series model was ARIMA (2, 0, 1) × (0, 0, 1)12 which fitted the SFTS cases reasonably well during the training process and forecast process. In the spatial clustering analysis, the global autocorrelation suggested that SFTS cases were not of random distribution. Local spatial autocorrelation analysis of SFTS identified foci mainly concentrated in Hubei Province, Henan Province, Anhui Province, Shandong Province, Liaoning Province, and Zhejiang Province. A most likely cluster including 21 counties in Henan Province and Hubei Province was observed in the central region of China from April 2015 to August 2016. Our results will provide a sound evidence base for future prevention and control programs of SFTS such as allocation of the health resources, surveillance in high-risk regions, health education, improvement of diagnosis and so on.

Ecological Pressure of Carbon Footprint in Passenger Transport: Spatio-Temporal Changes and Regional Disparities

Passenger transport has become a significant producer of carbon emissions in China, thus strongly contributing to climate change. In this paper, we first propose a model of ecological pressure of the carbon footprint in passenger transport (EPcfpt). In the model, the EPcfpt values of all the provinces and autonomous regions of China are calculated and analyzed during the period of 2006–2015. For the outlier EPcfpt values of Beijing, Shanghai and Tianjin, the research areas are classified into two scenarios: the first scenario (all the provinces and autonomous regions) and the second scenario (not including Beijing, Shanghai and Tianjin). The global spatial autocorrelation analysis of the first scenario shows that the EPcfpt might be randomly distributed, while it shows positive spatial autocorrelation in the second scenario. Furthermore, we carry out the local spatial autocorrelation analysis of the second scenario, and find that the low aggregation areas are the most common type and are mainly located in the west of China. Then the disparities in EPcfpt between China’s Eight Comprehensive Economic Zones are further analyzed. Finally, we put forward a number of policy recommendations in relation to the spatio-temporal changes and the regional disparities of EPcfpt in China. This study provides related references for proposing effective policy measures to reduce the ecological pressure of carbon emissions from the passenger transport sector.

Dynamic Assessment on the Landscape Patterns and Spatio-temporal Change in the mainstream of Tarim River

The Tarim River (TR), as the longest inland river at an arid area in China, is a typical regions of vegetation variation research and plays a crucial role in the sustainable development of regional ecological environment. In this paper, the newest dataset of MODND1M NDVI, at a resolution of 500m, were applied to calculate vegetation index in growing season during the period 2000-2015. Using a vegetation coverage index, a trend line analysis, and the local spatial autocorrelation analysis, this paper investigated the landscape patterns and spatio-temporal variation of vegetation coverage at regional and pixel scales over mainstream of the Tarim River, Xinjiang. The results showed that (1) The bare land area on both sides of Tarim River appeared to have a fluctuated downward trend and there were two obvious valley values in 2005 and 2012. (2) Spatially, the vegetation coverage improved areas is mostly distributed in upstream and the degraded areas is mainly distributed in the left bank of midstream and the end of Tarim River during 2000-2005. (3) The local spatial auto-correlation analysis revealed that vegetation coverage was spatially positive autocorrelated and spatial concentrated. The high-high self-related areas are mainly distributed in upstream, where vegetation cover are relatively good, and the low-low self-related areas are mostly with lower vegetation cover in the lower reaches of Tarim River.

Local Spatial Autocorrelation Analysis of 3 Disease Prevalence: A Case Study of Korea

Local Spatial Autocorrelation Analysis of 3 Disease Prevalence: A Case Study of Korea

Epidemiologic and spatio-temporal characteristics of hepatitis E in China, 2004-2014

Objective: To describe and analyze the epidemiologic and spatio-temporal characteristics of hepatitis E in China from 2004 to 2014. Methods: Data on the incidence of hepatitis E in 31 provinces (municipality and autonomous region) from 2004 to 2014, were collected. Empirical Mode Decomposition (EMD) was applied to decompose the time-series data to accurately describe the trend of hepatitis E incidence. Mathematic model was used to estimate the annual change of incidence in each age group and the whole province. Software ArcGIS 10.1 and SaTScan 9.01 were used to analyze the spatio-temporal clusters. Results: During 2004-2014, a total of 245 414 hepatitis E cases were reported in China. The overall incidence showed a slight increase (OR=1.05, 95%CI: 1.03-1.10). Incidence rates on hepatitis E were discovered different across the provinces, with significant increase appearing in the southern, central and northwestern areas. The highest increase was seen in the elderly, especially in the 65-69 and 70-74 year-olds. Results from the Local spatial autocorrelation analysis showed that the "high-high cluster" was moving from the north to the south and the "low-low cluster" disappeared as time went by. Data from Spatio-temporal scanning showed that there were five spatio-temporal clustering areas across the country. Conclusion: The overall incidence of hepatitis E was on the rise from 2004 to 2014, in China, but with differences seen across the areas and age groups.

Analysis of spatiotemporal patterns and influential factors for mushroom poisoning in Hunan Province in 2015

To understand the spatial distribution of mushroom poisoning in Hunan Province and its influential factors, and to provide the evidence for control of the mushroom poisoning. Methods: The surveillance data for mushroom poisoning cases from 122 counties in Hunan Province in 2015 were collected. Based on geographical information system database, spatial autocorrelation analysis and spatial regression analysis (via OpenGeoDa) was conducted. Results: The incidence of mushroom poisoning in Hunan Province in 2015 was 2.94/100 000. Global Moran's I values was 0.315 (P<0.05). Local spatial autocorrelation analysis indicated that Ningxiang, Xiangtan, Shaoyang, Lingling, Jiahe, and Linwu districts et al were "positive hotspot" regions. Guzhang, Hecheng, Dingcheng, Yueyang districts et al were "negative hotspot" regions. Spatial regression analysis revealed that the reported incidence of mushroom poisoning was positively correlated with the annual average temperate (Z=2.145, P=0.032), the number of health care institutions per capita (Z=2.352, P=0.019), and the number of students enrollment in secondary schools (Z=4.309, P<0.001). It was negatively associated with the number of school staff and workers of secondary schools (Z=-2.626, P=0.009). Conclusion: The spatial distribution of mushroom poisoning cases in Hunan Province in 2015 is highly clustered. Mushroom poisoning cases are more prevalent in the middle and southern regions and less prevalent in the northern areas of the province. The annual average temperate and the number of students enrollment in secondary schools demonstrate a certain positive influence on the distribution of mushroom poisoning in Hunan.

Analysis on epidemiology and spatial-temporal clustering of human brucellosis in Fujian province, 2011-2016

Objective: To analyze the epidemiological characteristics and spatial distribution of human brucellosis in Fujian province during 2011-2016, and provide evidence for the prevention and control of the disease. Methods: The surveillance data of human brucellosis in Fujian during 2011-2016 was analyzed with software R 3.3.1, ArcGIS 10.3.1, GeoDa 1.8.8 and SaTScan 9.4.3. Results: During 2011-2016, a total of 319 human brucellosis cases were reported, the incidence increased year by year (F=11.838, P=0.026) with the annual incidence of 0.14/100 000. The male to female rate ratio of the incidence was 2.50 ∶ 1. Farmers and herdsmen accounted for 57.37%. The incidence was 0.40/100 000 in Zhangzhou and 0.32/100 000 in Nanping, which were higher than other areas. The number of affected counties (district) increased from 12 in 2011 to 28 in 2016, showing a significant increase (F=13.447, P=0.021). The Moran's I of brucellosis in Fujian between January 2011 and December 2016 was 0.045, indicating the presence of a high value or low value clustering areas. Local spatial autocorrelation analysis showed that, high-high clustering area (hot spots) were distributed in Zhangpu, Longhai, Longwen, etc, while high-low clustering areas were distributed in Nan'an and Jiaocheng, etc. Temporal scanning showed that there were three clustering areas in areas with high incidence, the most possible clustering, occurring during January 1, 2013- December 31,2015, covered 6 counties, including Yunxiao, Pinghe, Longhai, etc, and Zhangpu was the center, (RR=7.96, LLR=92.62, P<0.001). Conclusions: The epidemic of human brucellosis in Fujian is becoming serious, and has spread to general population and non-epidemic areas. It is necessary to strengthen the prevention and control of human brucellosis in areas at high risk.

Epidemiology characteristics of crawfish related rhabdomyolysis in Nanjing, 2016: a multicenter retrospective investigation

To investigate the epidemiology characteristics of crawfish related rhabdomyolysis (RM) in Nanjing, 2016. Outpatient and inpatient electronic medical system of 21 hospitals in Nanjing during 2016 were retrospectively searched, and all the patients diagnosed with RM were selected. The patients with none crayfish-related RM was excluded. The epidemiology characteristics were depicted. The geographic information system (GIS) was used to collect, manage and analyze the spatial data, to visualize it, to analyze the spatial distribution features of the disease, and to explore the cause of disease prediction. GeoDa 1.8 software was used to analyze the global and local spatial auto-correlation. A total of 1 183 patients with crawfish related RM were initially screened, excluding 59 patients with RM caused by trauma, severe exercise, heat stroke, myositis, poisoning, drugs, and genetic diseases, and 1 124 patients were enrolled. The proportion of men was 36.48% (410/1 124) with an incidence of 12.54/100 thousands; while of women was 63.52% (714/1 124) with an incidence of 21.86/100 thousands. The median age at onset was 34 (28, 43) years. From July to August, the incidence of crawfish related RM was the highest, accounting for 96.53% of the total number of cases. The top four incidence areas were Pukou (41.54/100 thousands), Jianye (25.94/100 thousands), Qixia (25.73/100 thousands), Gulou (25.04/100 thousands), all of which were adjacent to the Yangtze River. Global spatial autocorrelation analysis showed: Moran I = 0.427, Z = 2.646, P = 0.003, suggesting that the crawfish related RM had positive spatial autocorrelation. The results showed that the spatial structure of crawfish related RM existed in Nanjing in 2016. Local spatial autocorrelation analysis showed that the "high-high" concentration areas were Pukou, Jianye and Liuhe. The incidences of above three areas which were the Nanjing section of the lower reaches of the Yangtze River flowed through the region and surrounding areas were higher than the overall incidence of Nanjing. The prevalence of crawfish related RM in Nanjing during 2016 had an obvious region-concentrated character and global spatial autocorrelation with the high prevalent regions mainly concentrated in the urban areas adjacent to the Yangtze River.

Life expectancy impacts due to heating energy utilization in China: Distribution, relations, and policy implications

The relation between life expectancy and energy utilization is of particular concern. Different viewpoints concerned the health impacts of heating policy in China. However, it is still obscure that what kind of heating energy or what pattern of heating methods is the most related with the difference of life expectancies in China. The aim of this paper is to comprehensively investigate the spatial relations between life expectancy at birth (LEB) and different heating energy utilization in China by using spatial autocorrelation models including global spatial autocorrelation, local spatial autocorrelation and hot spot analysis. The results showed that: (1) Most of heating energy exhibit a distinct north-south difference, such as central heating supply, stalks and domestic coal. Whereas spatial distribution of domestic natural gas and electricity exhibited west-east differences. (2) Consumption of central heating, stalks and domestic coal show obvious spatial dependence. Whereas firewood, natural gas and electricity did not show significant spatial autocorrelation. It exhibited an extinct south-north difference of heat supply, stalks and domestic coal which were identified to show significant positive spatial autocorrelation. (3) Central heating, residential boilers and natural gas did not show any significant correlations with LEB. While, the utilization of domestic coal and biomass showed significant negative correlations with LEB, and household electricity shows positive correlations. The utilization of domestic coal in China showed a negative effect on LEB, rather than central heating. To improve the solid fuel stoves and control consumption of domestic coal consumption and other low quality solid fuel is imperative to improve the public health level in China in the future.

The Effect of Urbanization and Farmland Transfer on the Spatial Patterns of Non-Grain Farmland in China

Rapid urbanization has accelerated farmland transfer in China, which in turn has increased the cultivation intensity of non-grain farmland (NGF). The aim of this paper was to analyze the spatial patterns of non-grain farmland (NGF) at municipal-level cities in China from 2000 to 2014, and explore the influence of urbanization and farmland transfer on the spatial patterns of non-grain farmland (NGF). The study found that both NGF area and share have a consistent variation trend which increased first, then decreased, and increased again. There were great differences between NGF area and share, which had various spatial patterns. Global Moran’s I of NGF area and share had strong positive spatial autocorrelations in 2000, 2005, 2010, and 2014, and revealed remarkable clustered spatial patterns. Local spatial autocorrelation analysis found that the spatial agglomeration pattern of the NGF area and the share changed significantly from 2000 to 2014. The spatial patterns of NGF were significantly influenced by regional differences in urbanization and farmland transfer. These results provide critical insights for formulating differential farmland transfer policies and policies for sustainable development of agriculture and rural economy in China.

Geographical Environment Factors and Risk Assessment of Tick-Borne Encephalitis in Hulunbuir, Northeastern China.

Tick-borne encephalitis (TBE) is one of natural foci diseases transmitted by ticks. Its distribution and transmission are closely related to geographic and environmental factors. Identification of environmental determinates of TBE is of great importance to understanding the general distribution of existing and potential TBE natural foci. Hulunbuir, one of the most severe endemic areas of the disease, is selected as the study area. Statistical analysis, global and local spatial autocorrelation analysis, and regression methods were applied to detect the spatiotemporal characteristics, compare the impact degree of associated factors, and model the risk distribution using the heterogeneity. The statistical analysis of gridded geographic and environmental factors and TBE incidence show that the TBE patients mainly occurred during spring and summer and that there is a significant positive spatial autocorrelation between the distribution of TBE cases and environmental characteristics. The impact degree of these factors on TBE risks has the following descending order: temperature, relative humidity, vegetation coverage, precipitation and topography. A high-risk area with a triangle shape was determined in the central part of Hulunbuir; the low-risk area is located in the two belts next to the outside edge of the central triangle. The TBE risk distribution revealed that the impact of the geographic factors changed depending on the heterogeneity.

Analysis of Spatial Clustering of HIV infected in men who have sex with men in Chongqing of 2004-2015

Receiving voluntary counseling and testing in the urban area of Chongqing and willing to participate in the interview by oral informed consent; male and self-described as MSM. The content of the interview included basic information, sexual orientation, sexual role, the main place of making friends, the main place of sexual behavior, a long-term experience in other provinces and drug abuse. Results: The HIV/AIDS reported number in Chongqing from 2004 to 2015 showed an uptrend, except in 2010. The age distribution of 6 604 cases of HIV positive patients was mainly concentrated in the 15-34 years old, about 68.5% (4 522 cases). There was a positive spatial autocorrelation in MSM, except 2005 (Moran's I=-0.046, P=0.823), form 2004 to 2015, Global Moran's I values were 0.308, 0.254, 0.335, 0.683, 0.673, 0.558, 0.620, 0.673, 0.685, 0.654 and 0.649, respectively; all P values were <0.01. The result of local spatial autocorrelation analysis showed that high-high accumulation area development in Chongqing city was divided into two stages in 2004-2015, which were Yuzhong, Jiangbei, and Shapingba district in 2004-2007 and the expanded Jiulongpo, Nan'an, and Yubei district in 2008-2015. Qualitative interviews results revealed that the age of the respondents was 20-44, and the mainly way of making friends were using mobile phone App and internet (17 participants). Most of the participants (11 participants) were making friends in the bar. The majority of respondents would ask the friends or themselves to use condoms when meeting with them the first time (19 participants), and 8 of respondents reported that they would not use condoms when their old friends refused to use condoms. Conclusion: The HIV infected MSM mainly aged between 15-34 years old and the spatial distribution of HIV/AIDS among MSM was clustered in economically developed main area in Chongqing. MSM began to make friends on the Internet, and could not adhere to using condom, which indicated that we should focus on internet intervention to find more efficacious interventions.

Evolution of landslide hotspots in Taiwan

Among the disasters facing Taiwan, earthquakes and typhoons incur the greatest monetary losses, and landslide disasters inflict the greatest damage in mountainous areas. The nationwide landslide susceptibility map gives an indication of where landslides are likely to occur in the future; however, there is no objective index indicating the location of landslide hotspots. In this study, we used statistical analysis to locate landslide hotspots in catchments in Taiwan. Global and local spatial autocorrelation analysis revealed the existence of landslide clusters between 2003 and 2012 and identified a concentration of landslide hotspots in the eastern part of Central Taiwan. The extreme rainfall brought by typhoon Morakot also led to the formation of new landslide hotspots in Southern Taiwan. This study provides a valuable reference explaining changes in landslide hotspots and identifying areas of high hotspot concentration to facilitate the formulation of strategies to deal with landslide risk.

Spatial distribution of potential and positive Aedes aegypti breeding sites

The spatial distribution of Aedes aegypti is heterogeneous, and the interaction between positive and potential breeding sites located both inside and outside homes is one of the most difficult aspects to characterize in vector control programs. To describe the spatial relationship between potential and positive breeding sites of A. aegypti inside and outside homes in Cali, Colombia. We conducted an entomological survey to collect data from both indoor and outdoor breeding sites. The exploratory analysis of spatial data included location, spatial trends, local spatial autocorrelation, spatial continuity and spatial correlation of positive and potential breeding sites according to habitat. Spatial trends were identified, as well as clusters of potential and positive breeding sites outdoors using local spatial autocorrelation analysis. A positive correlation was found between potential and positive breeding sites, and a negative correlation existed between indoor and outdoor sites. The spatial relationship between positive and potential A. aegypti breeding sites both indoors and outdoors is dynamic and highly sensitive to the characteristics of each territory. Knowing how positive and potential breeding sites are distributed contributes to the prioritization of resources and actions in vector control programs.

Distribución espacial de criaderos positivos y potenciales de Aedes aegypti

Introduction: The spatial distribution of Aedes aegypti is heterogeneous, and the interaction between positive and potential breeding sites located both inside and outside homes is one of the most difficult aspects to characterize in vector control programs. Objective: To describe the spatial relationship between potential and positive breeding sites of A. aegypti inside and outside homes in Cali, Colombia. Materials and methods: We conducted an entomological survey to collect data from both indoor and outdoor breeding sites. The exploratory analysis of spatial data included location, spatial trends, local spatial autocorrelation, spatial continuity and spatial correlation of positive and potential breeding sites according to habitat. Results: Spatial trends were identified, as well as clusters of potential and positive breeding sites outdoors using local spatial autocorrelation analysis. A positive correlation was found between potential and positive breeding sites, and a negative correlation existed between indoor and outdoor sites. Conclusions: The spatial relationship between positive and potential A. aegypti breeding sites both indoors and outdoors is dynamic and highly sensitive to the characteristics of each territory. Knowing how positive and potential breeding sites are distributed contributes to the prioritization of resources and actions in vector control programs.

Autocorrelation analysis of pneumoconiosis in China

Objective: Using spatial autocorrelation analysis to study spatial distribution characteristics of pneumoconiosis in China. Methods: The incidence of pneumoconiosis of each province from 2009 to 2013 was calculated by reported data of pneumoconiosis and the number of employment in mining and manufacturing of each province from Chinese statistical yearbook. The global and local spatial autocorrelation analysis was im-plemented by global and local's Moran's I index respectively. Results: A total of 112 066 new cases of pneumo-coniosis were reported from 2009 to 2013 in China, Global Moran's I from 2009 to 2013 was -0.114、0.179、0.199、0.259 and 0.241 respectively. A total of 15 space units with statistical significance were detected by local spatial autocorrelation analysis, which were mainly high-high regions. Conclusion: There was weak global spa-tial autocorrelation relationship of the incidence of pneumoconiosis from 2010 to 2013 among the provinces in China. The results of local spatial autocorrelation analysis suggested that there were clustering regions of pneumo-coniosis in the southwest and central south of China.

Driving forces and the spatial patterns of industrial sulfur dioxide discharge in China

Rapid industrialization in China has brought forward serious and harmful atmospheric pollution. In this study, spatial econometric analysis was used to analyze the spatial change and the driving forces behind the industrial sulfur dioxide (SO2) discharge in China from 2001 to 2014. The study found that the amount of industrial SO2 discharge (ISOD) increased first and then decreased during this period. ISOD intensity witnessed a fluctuating drop. There were large differences among intercity ISOD amount and intensity, which had various spatial patterns. Global Moran's I of ISOD amount and intensity had a tendency to increase on the whole, showed positive spatial autocorrelation, and revealed a more and more remarkable clustered spatial pattern. Local spatial autocorrelation analysis found that the spatial patterns of ISOD amount and intensity changed considerably over space and time. The spatial patterns of ISOD were significantly influenced by the regional differences in land use and environmental policy. The study also found that the driving forces of ISOD in China changed significantly from 2001 to 2014.