Quantifying bias in pattern indices extracted from spatially offset landscape samples

Landscape pattern is a mixture of natural and human-managed patches that vary in size, shape, arrangement and is the result of complex interactions of physical, biological, and social forces. At present, landscape pattern indices (LPIs) have become increasingly popular for characterizing landscape pattern and they are commonly calculated from land cover maps extracted from remotely sensed imagery. Pixel-based hard classification methods assign each pixel to one class and always suffer from the mixed pixel problem, and make the resultant land cover maps and extracted LPIs inaccuracy. Moreover, as the pixel size of hard classification maps is identical with the resolution of the remotely sensed imagery, spatial patterns of patches which are smaller than the resolution of remotely sensed imagery are usually ignored, and hard classification maps thus can hardly represent the spatial pattern of finer scale land cover classes due to the grain size effect. In general, soft classification techniques can estimate the class composition of land cover classes, however, their output fraction images provide no indication of how such classes are spatially distributed. Sub-pixel mapping (SPM) is a technology to transform the resulting fraction images of soft classification technologies into a finer scale hard classification map, which can solve the mixed pixel problem and eliminate the influence of grain size effect to a certain extent while providing the spatial representation of land cover targets at the sub-pixel scale. In this research, 33 landscape-level LPIs (for all land cover classes altogether) and 28 class-level LPIs (for individual land cover class) were applied to assess the performance of SPM method in characterizing landscape pattern. Ikonos imagery of Dujiangyan (Sichuan Province, China) representing urban landscape pattern and Landsat TM imagery of suburb of Wuhan (Hubei Province, China) representing rural landscape pattern were used in this study. We derived LPIs from both SPM maps and hard classification maps and found that SPM can better depict landscape pattern than hard classification methods at both the landscape level and class level in most cases.

The spatial characteristics and relationships between landscape pattern and ecosystem service value along an urban-rural gradient in Xi’an city, China

Urbanization has substantially changed landscape patterns and seriously disturbed the structure and function of the ecosystems. However, the spatial characteristics and relationships between landscape patterns and ecosystem service values (ESVs) along the urban-rural gradient remain unclear. Based on high-resolution images, this study used concentric buffer zones to explore the characteristics and relationship between landscape pattern indexes (LPIs) and ESVs in the rural-urban gradient to reveal the impact of urban development on urban ecosystems. The results showed that the landscape heterogeneity was high in the urban fringe 18-20 km from the urban center. The PD, ED, LSI, SHAPE_MN, DIVISION, SPLIT, and SHIDI variables had the lowest values in the urban center, while CONTAG and AI had the high values in the urban. Water bodies and forest land are the main land use/land cover (LULC) types that provide ecosystem services. The total ESV of Kunshan city totaled 5597.31 × 106 CNY in 2018. The average ESV increased from 2.42 × 106 CNY to 9.92 × 106 CNY along the urban-rural gradient, which indicated that natural landscapes had higher ESVs. ED and Landscape Division Index (DIVISION) had positive effects on ESV, while Largest Patch Index (LPI), Contagion (CONTAG), Proportion of Like Adjacencies (PLADJ), Patch Cohesion Index (COHESION), and Aggregation Index (AI) had negative effects on ESV. The results of the regression model indicated there were quantitative relationships between ESVs and LPIs, which revealed how landscape pattern affected ESVs. The study can provide a scientific reference for the optimization of urban landscape patterns and urban and rural sustainable development.

The extent to which landscape spatial patterns can impact the dynamics and distribution of biodiversity is a key geography and ecology issue. However, few previous studies have quantitatively analyzed the spatial relationship between the landscape pattern and habitat quality from a simulation perspective. In this study, the landscape pattern in 2031 was simulated using a patch-generating simulation (PLUS) model for the Yellow River Basin. Then, the landscape pattern index and habitat quality from 2005 to 2031 were evaluated using the Fragstats 4.2 and the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model. Furthermore, we analyzed the spatial distribution characteristics and spatial spillover effects of habitat quality using spatial autocorrelation analysis. Finally, the spatial association between the landscape pattern index and habitat quality was quantitatively revealed based on a spatial lag model. The simulation results showed that: (1) from 2005 to 2031, the landscape of the Yellow River Basin would be dominated by grassland and unused land, and the areas of construction land and water body will increase significantly, while the area of grassland will decrease; (2) patch density (PD) and Shannon’s diversity index (SHDI) show significant increases, while edge density (ED), landscape shape index (LSI), mean patch area (AREA_MN), and contagion index (CONTAG) decrease; (3) from 2005 to 2031, habitat quality would decrease. The high-value areas of habitat quality are mainly distributed in the upper reaches of the Yellow River Basin, and the low-value areas are distributed in the lower reaches. Meanwhile, both habitat quality and its change rate present positive spatial autocorrelation; and (4) the spatial relationships of habitat quality with PD and COHESION are negative, while ED and LSI have positive impacts on habitat quality. Specifically, landscape fragmentation caused by high PD has a dominant negative influence on habitat quality. Therefore, this study can help decision makers manage future landscape patterns and develop ecological conservation policy in the Yellow River Basin.

In mountainous areas, the landscape spatial pattern is affected by topography and human activities. To better understand topographic influences on the vegetation structure in the landscape mosaic and interpret the ecological implications of spatial patterns, the relationships between topographic features and landscape pattern indices (LPIs) should be addressed. The objective of this study is to address the effects of topography on LPIs (both planimetric and surface LPIs) by integrating spatial analysis and gradient analysis. To fulfill these objectives, two different mountain research sites located in the Lancang River watershed were chosen, both planimetric and surface landscape metrics were used to quantify natural category (NC) and anthropogenic category (AC) patterns, and the redundancy ordination gradient method was used to address the relationship between LPIs (planimetric and surface metrics) of NC and AC patterns and topographic attributes. The results suggest that topographic attributes play a prominent role in structuring the landscape mosaic. Elevation is a key factor for NC and AC patch patterns in the study area. The results indicate that the fragmentation and isolation of NC patterns increase as elevation decreases. In contrast, the fragmentation and isolation of AC patterns decrease as elevation decreases.

Comprehending the dynamic change characteristics of land use/cover and the driving factors causing the change are prerequisites for protecting land resources. This paper analyzes changes in cultivated land, the driving factors that cause them, and their tremendous impact on landscape pattern changes in the Dongting Lake Basin. For this purpose, we used mathematical statistics, buffer analysis, trend analysis, landscape pattern index, and logistic regression model to analyze the land use data of the study area from 1980 to 2018. The results show that the cultivated land showed a decreasing trend, with the total area decreased by 4.76% (or 716.13 km2) from 1980 to 2018, and the activity of mutual transformation with other land use types decreased. The spatial distribution pattern of cultivated land and landscape shows the change characteristics gradually from Dongting Lake to the surroundings. Among the driving factors of cultivated land changes, the influence of human activities was gradually increasing, while the natural factors were decreasing. The cultivated land landscape pattern index and the overall landscape pattern index have a significant positive correlation, showing relatively consistent change trend and spatial distribution characteristics. We believe that the decrease of cultivated land area has a certain relationship with the increase of landscape fragmentation in the Dongting Lake Basin. Our research is expected to provide a reference for strengthening regional cultivated land management and rational development and utilization of regional land resources.

In the application of remote sensing it is common to investigate processes that generate patches of material. This is especially true when using categorical land cover or land use maps. Here we view some existing tools, landscape pattern indices (LPI), as non-parametric estimators of random closed sets (RACS). This RACS framework enables LPIs to be studied rigorously. A RACS is any random process that generates a closed set, which encompasses any processes that result in binary (two-class) land cover maps. RACS theory, and methods in the underlying field of stochastic geometry, are particularly well suited to high-resolution remote sensing where objects extend across tens of pixels, and the shapes and orientations of patches are symptomatic of underlying processes. For some LPI this field already contains variance information and border correction techniques. After introducing RACS theory we discuss the core area LPI in detail. It is closely related to the spherical contact distribution leading to conditional variants, a new version of contagion, variance information and multiple border-corrected estimators. We demonstrate some of these findings on high resolution tree canopy data.

The study on landscape pattern is an important field of urban land use and ecological change. Since 1990s, the widely accepted Patch-Corridor-Matrix model is generally used in qualitative description of landscape pattern. In recent years, quantitative evaluation on urban landscape dynamics is becoming hot in research. By making a critical review on existing research methods of landscape pattern, a new approach based on RS/GIS is put forward in this paper, comprising three steps, pattern characteristics - Gradient differentiation feature- Directional signature of the and we call it GGD. This method is applied to the case study of Xi'an metropolitan area in China. The result shows that the method is effective on quantitative study of urban landscape. The preparation of the method GGD is setting up research platform based on RS and GIS. By using the software of Geographical Information System (Arcgis9.0 & Erdas), the authors got the interpretation of remote sensing images of different years, and carried on the division of the landscape type of the research region. By calculating various index of landscape level with software Fragstats3.3 as an assistant tool and adopting three steps of GGD combined with landscape index, this paper can assesses the landscape spatial pattern of urban area: 1) General pattern characteristics analysis is to get transition probability of various landscape through Markov chain and to predict the landscape transformation by introducing CA model. The analysis emphasizes on total landscape structure and its change over time; 2) Gradient characteristic analysis, which makes gradient zone by taking city as a center outwardly with certain distance and contrastively analyzes the landscape index of each subarea, stresses the spatial character of landscape pattern, verifies urban morphology theories and provides the quantitative warranty for establishment of urban modality. Therefore, the analysis is useful for supervising urban expanding speed; 3)Direction characteristic analysis, which is setting up radiate strip on west-east, south-north, southwest-northeast and northwest-southeast and form certain width on each direction, can precisely and quantitatively indicate different characteristic of urban landscape at each development direction, and by combined with gradient analysis it is highly advantageous to the examination and planning of urban expanding direction. In the case study on Xi'an metropolitan area, remote sensing images of 1988 and 2005 Landsat-TM were handled, and the division of the landscape type of the region was also carried on. According to the above approach, the result was got and some valuable information was showed as follows: 1) The diversity of overall landscape of Xi'an metropolitan area tends to increase and the degree of fragmentation tends to deepen. With the increase of urbanization level, the visual component of landscape is more and more diversified, the shapes of landscape is more and more complicated and ecologically more and more fragmented. In the region where urbanization level is low, natural landscape is the main component of the landscape, the diversity of the landscape is low. And because landscape is seriously disturbed by human activities with urbanization, fragmentation of the landscape emerges periodically. 2) In the process of transect gradient analysis, the landscape pattern index can explore the urbanization gradient, and its trend to reduce gradually towards the suburban. The landscape of area with a high urbanization level is mainly man-created, and its patches show large number, small area, simple shape and higher landscape heterogeneity. The transect gradient analysis on different time series indicates the relationship between urbanization level and landscape pattern. The landscape of urban area suffers intensely from human being, and its pattern component and spatial collocation depends on the interference intensity to a large degree. In the area with a high urbanization level, its pattern component is more man-created and less natural landscape. The landscape collocation characteristic of its patches takes on a large number, little average area, simple shape and low polymerization degree. 3) Analysis of direction and gradient of Xi'an metropolitan area can quantitatively reflect influence of urbanization and characteristics of urban landscape in the main development axes of north-south and east. Result shows that the degree of internal integration between Xi'an city and Xianyang city is gradually enhanced with the quickly urbanization course in China.

Influence of landscape patterns on nitrate and particulate organic nitrogen inputs to urban stormwater runoff

Landscape pattern and ecosystem services are critical for protected areas' contributions to sustainable development goals at regional scale

The correlation analysis on the landscape pattern index and hydrological processes in the Yanhe watershed, China

Urbanization has become an important contributor for global warming and urban air temperature is rising gradually in all the cities. It means that urban land use and land cover (LULC) changed became critical in determining the urban environment quality. This paper presents an integrated study to investigate and identify landscape pattern which have the influence to increase of land surface temperature, using Geographical Information Systems (GIS) and remote sensing. In this paper, an oasis city-Urumqi as an example, which explores the temporal and spatial characteristics of land use and land cover change from 1990 to 2004, and the land surface temperature variation from 1990 to 2004. Results shows that land use of Urumqi appeared a dramatic variation during the last 15 years. The mean annul temperatures of the land surface in Urumqi were 21.41 °C in 1990 and 31.12 °C in 2004, respectively. In particular, the land surface temperature and landscape pattern index had shown the close relationship. To sum up, integration of remote sensing and GIS was an effective way to monitor and analyze urban LULC, and evaluate the landscape pattern impact on land surface temperature.

In order to overcome the problems of large prediction error and slow prediction speed in dynamic prediction method of urban landscape, a new dynamic prediction method of urban landscape pattern based on remote sensing image fusion is proposed. The high-resolution remote sensing image is preprocessed by atmospheric correction and radiometric correction to obtain the fusion result of high-resolution remote sensing image. The urban landscape information is extracted from the fusion results, and the corresponding urban landscape pattern analysis model is built. The landscape pattern index is set up in the model, and the dynamic prediction of urban landscape pattern is completed by combining the characteristics of index change with the driving force and image factors of urban pattern change. The experimental results show that the average prediction error is 1.56, and the operation delay is reduced by 2 s-3 s.

The landscape index is a quantitative reflection of different landscape structural components and spatial pattern characteristics through highly concentrated landscape pattern information. The landscape pattern index can be divided into three levels: Patch metrics, Class metrics, and Landscape metrics, among which the patch level index is the basis for calculating other landscape indices, reflecting the landscape pattern by calculating the area, perimeter, number, shape and other characteristics of each patch in the landscape. There are differences in the landscape indices of different families of plants, the range of plant species involved in our study is small, and future studies can add different plant species as the dominant grassland communities. Also in future studies, the ecological implications of small-scale patterns should be explored under different vegetation cover, climate, soil, topography, and vegetation types.

Analyzing the dynamic changes and causes of greenspace landscape patterns in Beijing plains