Assessing uncertainties in land cover projections

Understanding uncertainties in land cover projections is critical to investigating land‐based climate mitigation policies, assessing the potential of climate adaptation strategies and quantifying the impacts of land cover change on the climate system. Here, we identify and quantify uncertainties in global and European land cover projections over a diverse range of model types and scenarios, extending the analysis beyond the agro‐economic models included in previous comparisons. The results from 75 simulations over 18 models are analysed and show a large range in land cover area projections, with the highest variability occurring in future cropland areas. We demonstrate systematic differences in land cover areas associated with the characteristics of the modelling approach, which is at least as great as the differences attributed to the scenario variations. The results lead us to conclude that a higher degree of uncertainty exists in land use projections than currently included in climate or earth system projections. To account for land use uncertainty, it is recommended to use a diverse set of models and approaches when assessing the potential impacts of land cover change on future climate. Additionally, further work is needed to better understand the assumptions driving land use model results and reveal the causes of uncertainty in more depth, to help reduce model uncertainty and improve the projections of land cover.     

Effects of land use change and management on the European cropland carbon balance

We model the carbon balance of European croplands between 1901 and 2000 in response to land use and management changes. The process‐based ORCHIDEE‐STICS model is applied here in a spatially explicit framework. We reconstructed land cover changes, together with an idealized history of agro‐technology. These management parameters include the treatment of straw and stubble residues, application of mineral fertilizers, improvement of cultivar species and tillage. The model is integrated for wheat and maize during the period 1901–2000 forced by climate each 1/2‐hour, and by atmospheric CO₂, land cover change and agro‐technology each year. Several tests are performed to identify the most sensitive agro‐technological parameters that control the net biome productivity (NBP) in the 1990s, with NBP equaling for croplands the soil C balance. The current NBP is a small sink of 0.16 t C ha^?1 yr^?1. The value of NBP per unit area reflects past and current management, and to a minor extent the shrinking areas of arable land consecutive to abandonment during the 20th Century. The uncertainty associated with NBP is large, with a 1‐sigma error of 0.18 t C ha^?1 yr^?1 obtained from a qualitative, but comprehensive budget of various error terms. The NBP uncertainty is dominated by unknown historical agro‐technology changes (47%) and model structure (27%), with error in climate forcing playing a minor role. A major improvement to the framework would consist in using a larger number of representative crops. The uncertainty of historical land‐use change derived from three different reconstructions, has a surprisingly small effect on NBP (0.01 t C ha^?1 yr^?1) because cropland area remained stable during the past 20 years in all the tested land use forcing datasets. Regional cross‐validation of modeled NBP against soil C inventory measurements shows that our results are consistent with observations, within the uncertainties of both inventories and model. Our estimation of cropland NBP is however likely to be biased towards a sink, given that inventory data from different regions consistently indicate a small source whereas we model a small sink.     

Doubling protected land area may be inefficient at preserving the extent of undeveloped land and could cause substantial regional shifts in land use

Projection of land use and land-cover change is highly uncertain yet drives critical estimates of carbon emissions, climate change, and food and bioenergy production. We use new, spatially explicit land availability data in conjunction with a model sensitivity analysis to estimate the effects of additional land protection on land use and land cover. The land availability data include protected land and agricultural suitability and is incorporated into the Moirai land data system for initializing the Global Change Analysis Model. Overall, decreasing land availability is relatively inefficient at preserving undeveloped land while having considerable regional land-use impacts. Current amounts of protected area have little effect on land and crop production estimates, but including the spatial distribution of unsuitable (i.e., unavailable) land dramatically shifts bioenergy production from high northern latitudes to the rest of the world, compared with uniform availability. This highlights the importance of spatial heterogeneity in understanding and managing land change. Approximately doubling the current protected area to emulate a 30% protected area target may avoid land conversion by 2050 of less than half the newly protected extent while reducing bioenergy feedstock land by 10.4% and cropland and grazed pasture by over 3%. Regional bioenergy land may be reduced (increased) by up to 46% (36%), cropland reduced by up to 61%, pasture reduced by up to 100%, and harvested forest reduced by up to 35%. Only a few regions show notable gains in some undeveloped land types of up to 36%. Half of the regions can reach the target using only unsuitable land, which would minimize impacts on agriculture but may not meet conservation goals. Rather than focusing on an area target, a more robust approach may be to carefully select newly protected land to meet well-defined conservation goals while minimizing impacts to agriculture.     

Codominant water control on global interannual variability and trends in land surface phenology and greenness

Identifying the relative importance of climatic and other environmental controls on the interannual variability and trends in global land surface phenology and greenness is challenging. Firstly, quantifications of land surface phenology and greenness dynamics are impaired by differences between satellite data sets and phenology detection methods. Secondly, dynamic global vegetation models (DGVMs) that can be used to diagnose controls still reveal structural limitations and contrasting sensitivities to environmental drivers. Thus, we assessed the performance of a new developed phenology module within the LPJmL (Lund–Potsdam–Jena managed Lands) DGVM with a comprehensive ensemble of three satellite data sets of vegetation greenness and ten phenology detection methods, thereby thoroughly accounting for observational uncertainties. The improved and tested model allows us quantifying the relative importance of environmental controls on interannual variability and trends of land surface phenology and greenness at regional and global scales. We found that start of growing season interannual variability and trends are in addition to cold temperature mainly controlled by incoming radiation and water availability in temperate and boreal forests. Warming‐induced prolongations of the growing season in high latitudes are dampened by a limited availability of light. For peak greenness, interannual variability and trends are dominantly controlled by water availability and land‐use and land‐cover change (LULCC) in all regions. Stronger greening trends in boreal forests of Siberia than in North America are associated with a stronger increase in water availability from melting permafrost soils. Our findings emphasize that in addition to cold temperatures, water availability is a codominant control for start of growing season and peak greenness trends at the global scale.     

Global land cover characterization from satellite data: from research to operational implementation?

Satellite data provide the basis for geographically referenced global land cover characterization that is internally consistent, repeatable over time, and potentially more reliable than ground-based sources. During the last 20 years considerable research efforts have been devoted to the extraction of land cover information from these data. Only during the last few years have these methods begun to be applied in operational contexts. Such applications have thus far primarily addressed key global change issues such as the global carbon balance. Examples of the successful quasi-operational implementation of remote sensing include NASA's Humid Tropical Landsat Pathfinder project, where high resolution data are being used at subcontinental scales to measure change in the areal extent of tropical rain forests throughout the world, and the Tropical Ecosystem Environment observation by Satellite (TREES) project to assess forest cover in the tropics. At coarser resolutions, a number of land cover products suitable for incorporation in global and regional models have been developed. Alternatives to traditional land cover classifications have also been developed to describe gradients and mosaics in the vegetation more realistically. These land cover products offer the possibility for applications in ecological and human dimensions research at regional and global scales, as well as for implementation of international agreements that require land cover information. Recently launched and future satellites will carry sensors that provide data with greatly improved capabilities for land cover characterization and advancements in computing environments make it feasible to take advantage of these new data. However, several challenges must be overcome in making a transition from research to operational land cover monitoring, including automation of methods to analyse the satellite data, more effective techniques for validation, and assurance of long-term continuity in the availability of satellite measurements.     

A Land System representation for global assessments and land‐use modeling

Current global scale land‐change models used for integrated assessments and climate modeling are based on classifications of land cover. However, land‐use management intensity and livestock keeping are also important aspects of land use, and are an integrated part of land systems. This article aims to classify, map, and to characterize Land Systems (LS) at a global scale and analyze the spatial determinants of these systems. Besides proposing such a classification, the article tests if global assessments can be based on globally uniform allocation rules. Land cover, livestock, and agricultural intensity data are used to map LS using a hierarchical classification method. Logistic regressions are used to analyze variation in spatial determinants of LS. The analysis of the spatial determinants of LS indicates strong associations between LS and a range of socioeconomic and biophysical indicators of human‐environment interactions. The set of identified spatial determinants of a LS differs among regions and scales, especially for (mosaic) cropland systems, grassland systems with livestock, and settlements. (Semi‐)Natural LS have more similar spatial determinants across regions and scales. Using LS in global models is expected to result in a more accurate representation of land use capturing important aspects of land systems and land architecture: the variation in land cover and the link between land‐use intensity and landscape composition. Because the set of most important spatial determinants of LS varies among regions and scales, land‐change models that include the human drivers of land change are best parameterized at sub‐global level, where similar biophysical, socioeconomic and cultural conditions prevail in the specific regions.     

Climate and satellite-derived land cover for predicting breeding bird distribution in the Great Lakes Basin

Aim We examined relationships between breeding bird distribution of 10 forest songbirds in the Great Lakes Basin, large‐scale climate and the distribution of land cover types as estimated by advanced very high resolution radiometer (AVHRR) and multi‐spectral scanner (MSS) land cover classifications. Our objective was to examine the ability of regional climate, AVHRR (1 km resolution) land cover and MSS (200 m resolution) land cover to predict the distribution of breeding forest birds at the scale of the Great Lakes Basin and at the resolution of Breeding Bird Atlas data (5–10 km²). Specifically we addressed the following questions. (1) How well do AVHRR or MSS classifications capture the variation in distribution of bird species? (2) Is one land cover classification more useful than the other for predicting distribution? (3) How do models based on climate compare with models based on land cover? (4) Can the combination of both climate and land cover improve the predictive ability of these models. Location Modelling was conducted over the area of the Great Lakes Basin including parts of Ontario, Canada and parts of Illinois, Indiana, Michigan, New York, Ohio, Pennsylvania Wisconsin, and Minnesota, USA. Methods We conducted single variable logistic regression with the forest classes of AVHRR and MSS land cover using evidence of breeding as the response variable. We conducted multiple logistic regression with stepwise selection to select models from five sets of explanatory variables (AVHRR, MSS, climate, AVHRR + climate, MSS + climate). Results Generally, species were related to both AVHRR and MSS land cover types in the direction expected based on the known local habitat use of the species. Neither land cover classification appeared to produce consistently more intuitive results. Good models were generated using each of the explanatory data sets examined here. And at least one but usually all five variable sets produced acceptable or excellent models for each species. Main conclusions Both climate and large scale land cover were effective predictors of the distribution of the 10 forest bird species examined here. Models generated from these data had good classification accuracy of independent validation data. Good models were produced from all explanatory data sets or combinations suggesting that the distribution of climate, AVHRR land cover, and MSS land cover all captured similar variance in the distribution of the birds. It is difficult to separate the effects of climate and vegetation on the species’ distributions at this scale.     

Principles for life cycle inventories of land use on a global scale

Purpose

To assess the diverse environmental impacts of land use, a standardization of quantifying land use elementary flows is needed in life cycle assessment (LCA). The purpose of this paper is to propose how to standardize the land use classification and how to regionalize land use elementary flows.

Materials and methods

In life cycle inventories, land occupation and transformation are elementary flows providing relevant information on the type and location of land use for land use impact assessment. To find a suitable land use classification system for LCA, existing global land cover classification systems and global approaches to define biogeographical regions are reviewed.

Results and discussion

A new multi-level classification of land use is presented. It consists of four levels of detail ranging from very general global land cover classes to more refined categories and very specific categories indicating land use intensities. Regionalization is built on five levels, first distinguishing between terrestrial, freshwater, and marine biomes and further specifying climatic regions, specific biomes, ecoregions and finally indicating the exact geo-referenced information of land use. Current land use inventories and impact assessment methods do not always match and hinder a comprehensive assessment of land use impact. A standardized definition of land use types and geographic location helps to overcome this gap and provides the opportunity to test the optimal resolution of land cover types and regionalization for each impact pathway.

Conclusions and recommendation

The presented approach provides the necessary flexibility to providers of inventories and developers of impact assessment methods. To simplify inventories and impact assessment methods of land use, we need to find archetypical situations across impact pathways, land use types and regions, and aggregate inventory entries and methods accordingly.

     

The relationship between British ground beetles (Coleoptera, Carabidae) and land cover

Aims and methods Ground beetle and satellite‐derived land cover data from 1687 United Kingdom 10 km national grid squares were used to assess the relationship between species pool and cover data in Great Britain using fuzzy classification and constrained ordination. Results Ground beetle species pools classified into nine groups which were related to land cover variables using constrained ordination. There was a strong relationship between upland land cover and three ground beetle groups. Deciduous woodland, coastal and tilled land were associated with three other groups. Three further groups did not appear to be strongly associated with any particular cover, but differed in geographical position. Conclusion The distribution of species pools derived from the British national recording scheme at the 10 km scale was strongly related to satellite‐derived land cover data. There appears to be considerable potential for the use of a synthesis of land cover and ground beetle data in the monitoring of environmental change over a large, countrywide, area.     

基于长时间序列landsat数据的科尔沁沙地土地利用演变分析

以科尔沁沙地沙丘-草甸过渡带区域主要土地覆被类型为研究对象,以1987-2017年多时相Landsat TM/OLI遥感影像解译分类为基础,参考生态学植被演替研究方法,系统分析研究区30年来的土地利用/覆被动态演变规律,研究结果表明：（1）决策树法在复杂下垫面不同覆被类型的同步识别效果较好,所有影像分类精度均达到88%以上,分类效果较好,其中2017年分类精度最高为95.24%,达到了分类研究的要求;（2）研究区存在着"半灌丛-草甸地-灌丛"的植被结构特征,且整体表现为"南进北退"的变化趋势。结合土地利用动态度分析结果表明人类活动干涉下,研究区整体上遵循了半干旱区植被条件改善的一般规律,侧面反映该研究区域生态环境的持续不稳定性和脆弱性;（3）研究区覆被类型发生变化的总面积达到2623.59 hm²,总变化强度为63.76%。其中正向演替的比例为52.61%,以半灌丛面积的持续减小与沙地草甸面积的持续扩张为主要变化特征。但同时,半灌丛转为沙地的面积为184.95 hm²,表明以放牧为主的研究区同时发生着局部的逆行演变;（4）质心迁移结果反映了1987-2017年间,除人为影响较大的林地、草地以及耕地向北迁移外,其他植被类型的质心都有很明显的南迁,主要植被类型重心迁移距离依次由大到小为耕地 > 半灌丛 > 灌丛 > 沙地草甸 > 湿地草甸 > 林地。研究通过记录科尔沁沙地连续扩展的时空模式,展示了遥感-生态和时间序列影像在30 m分辨率下跟踪土地利用/覆被变化的潜力,为提高干旱半干旱区土地利用情况的动态监测效率,开展土地利用/覆被动态演变研究提供参考。     

Models meet data: Challenges and opportunities in implementing land management in Earth system models

As the applications of Earth system models (ESMs) move from general climate projections toward questions of mitigation and adaptation, the inclusion of land management practices in these models becomes crucial. We carried out a survey among modeling groups to show an evolution from models able only to deal with land‐cover change to more sophisticated approaches that allow also for the partial integration of land management changes. For the longer term a comprehensive land management representation can be anticipated for all major models. To guide the prioritization of implementation, we evaluate ten land management practices—forestry harvest, tree species selection, grazing and mowing harvest, crop harvest, crop species selection, irrigation, wetland drainage, fertilization, tillage, and fire—for (1) their importance on the Earth system, (2) the possibility of implementing them in state‐of‐the‐art ESMs, and (3) availability of required input data. Matching these criteria, we identify “low‐hanging fruits” for the inclusion in ESMs, such as basic implementations of crop and forestry harvest and fertilization. We also identify research requirements for specific communities to address the remaining land management practices. Data availability severely hampers modeling the most extensive land management practice, grazing and mowing harvest, and is a limiting factor for a comprehensive implementation of most other practices. Inadequate process understanding hampers even a basic assessment of crop species selection and tillage effects. The need for multiple advanced model structures will be the challenge for a comprehensive implementation of most practices but considerable synergy can be gained using the same structures for different practices. A continuous and closer collaboration of the modeling, Earth observation, and land system science communities is thus required to achieve the inclusion of land management in ESMs.     

Are CORINE land cover classes reliable proxies of plant species assemblages? A test in Mediterranean forest landscapes

In land cover mapping, the complexity of landscapes is fitted into classes that may limit the recognition of natural variability. In this study, we tested the power of land cover classes (defined on the CORINE land cover classification scheme, a standardized legend set by EU for land cover inventory) to separate different vascular plant assemblages in forest ecosystems. In order to separately identify the role of different sources of inconsistency between land cover classes and species composition, we compared three different inventory processes, based on (i) dominant tree species as observed in the field, (ii) visual interpretation of remotely sensed images and (iii) semi-automatic supervised classification of satellite images. Our results underline that classifying forest ecosystems on the basis of their canopy species produces an over-simplification of habitat variability. Consequently, land cover maps based on non-specialized classification schemes should not be regarded as good proxies for plant biodiversity. If land cover maps are intended to describe and manage landscapes and their associated biodiversity, it is necessary to improve their capacity to represent the complexity of ecosystems.     

National inventories of land occupation and transformation flows in the world for land use impact assessment

Purpose

Land use can cause significant impacts on ecosystems and natural resources. To assess these impacts using life cycle assessment (LCA) and ensure adequate decision-making, comprehensive national inventories of land occupation and transformation flows are required. Here, we aim at developing globally differentiated inventories of land use flows that can be used for primary use in life cycle impact assessment or national land planning.

Methods

Using publicly available data and inventory techniques, national inventories for several land use classes were developed. All land use classes were covered with the highest retrievable level of disaggregation within urban, forestry, agriculture and other land use classes, thus differentiating 21 land use classes. For illustrating the application of this newly developed inventory, two different application settings relevant to life cycle impact assessment were considered: the calculation of global normalisation references for 11 land use impact indicators related to soil quality assessment (adopting the methods recommended by the EU Commission) and the determination of generic globally applicable characterisation factors (CFs) resulting from aggregation of country-level CFs for situations for use when land use location is unknown.

Results and discussion

We built national inventories of 21 land occupation and 17 land transformation flows for 225 countries in the world for the reference year 2010. Cross-comparisons with existing inventories of narrower scopes attested its consistency. Detailed analyses of the calculated global normalisation references for the 11 land use impact categories showed different patterns across the land use impact indicators for each country, thus raising attention on key land use impacts specific to each country. Furthermore, the upscaling of country-level CFs to global generic CFs using the land use inventory revealed discrepancies with other alternative approaches using land use data at different resolutions.

Conclusions

In this study, we made a first attempt at developing national inventories of land use flows with sufficient disaggregation level to enable the calculation of normalisation references and differentiated impacts. However, the findings also demonstrated the need to refine the consistency of the inventory, particularly in the combination of land cover and land use data, which should be harmonised in future studies, and to expand it with differentiated coverage of more land use flows relevant to impact assessment.

     

Land Transitions in the Tropics: Going Beyond the Case Studies

Estimates of the percent of Earth's land surface that has either been transformed or degraded by human activity range between 39 and 50 percent, with agriculture accounting for the vast majority of these changes. Although much of the focus of research on land use and cover change in the tropics has been on deforestation, ongoing socioeconomic changes both locally and globally have made land transitions in the tropics extremely fluid. In addition, feedbacks between land cover change and human behavior constrain the extent and trajectories of land transitions. The sustainability of land use systems in the tropics depends on an understanding of coupled human–natural systems that can lead to general frameworks for management and prediction. The unprecedented availability of land use/cover data together with ecological data collected at large spatial scales offer exciting opportunities for advancing our understanding of socioecological systems. We rely on six studies of land transitions in the tropics to illustrate some promising approaches and pose critical questions to guide this body of research.     

Spatiotemporal land use modelling to assess land availability for energy crops – illustrated for Mozambique

A method and tool have been developed to assess future developments in land availability for bioenergy crops in a spatially explicit way, while taking into account both the developments in other land use functions, such as land for food, livestock and material production, and the uncertainties in the key determinant factors of land use change (LUC). This spatiotemporal LUC model is demonstrated with a case study on the developments in the land availability for bioenergy crops in Mozambique in the timeframe 2005–2030. The developments in the main drivers for agricultural land use, demand for food, animal products and materials were assessed, based on the projected developments in population, diet, GDP and self‐sufficiency ratio. Two scenarios were developed: a business‐as‐usual (BAU) scenario and a progressive scenario. Land allocation was based on land use class‐specific sets of suitability factors. The LUC dynamics were mapped on a 1 km² grid level for each individual year up to 2030. In the BAU scenario, 7.7 Mha and in the progressive scenario 16.4 Mha could become available for bioenergy crop production in 2030. Based on the Monte Carlo analysis, a 95% confidence interval of the amount of land available and the spatially explicit probability of available land was found. The bottom‐up approach, the number of dynamic land uses, the diverse portfolio of LUC drivers and suitability factors, and the possibility to model uncertainty mean that this model is a step forward in modelling land availability for bioenergy potentials.     

Critical land change information enhances the understanding of carbon balance in the United States

Large‐scale terrestrial carbon (C) estimating studies using methods such as atmospheric inversion, biogeochemical modeling, and field inventories have produced different results. The goal of this study was to integrate fine‐scale processes including land use and land cover change into a large‐scale ecosystem framework. We analyzed the terrestrial C budget of the conterminous United States from 1971 to 2015 at 1‐km resolution using an enhanced dynamic global vegetation model and comprehensive land cover change data. Effects of atmospheric CO₂ fertilization, nitrogen deposition, climate, wildland fire, harvest, and land use/land cover change (LUCC) were considered. We estimate annual C losses from cropland harvest, forest clearcut and thinning, fire, and LUCC were 436.8, 117.9, 10.5, and 10.4 TgC/year, respectively. C stored in ecosystems increased from 119,494 to 127,157 TgC between 1971 and 2015, indicating a mean annual net C sink of 170.3 TgC/year. Although ecosystem net primary production increased by approximately 12.3 TgC/year, most of it was offset by increased C loss from harvest and natural disturbance and increased ecosystem respiration related to forest aging. As a result, the strength of the overall ecosystem C sink did not increase over time. Our modeled results indicate the conterminous US C sink was about 30% smaller than previous modeling studies, but converged more closely with inventory data.     

Combining land cover mapping of coastal dunes with vegetation analysis

Question: Coastal dune systems are characterized by a natural mosaic that promotes species diversity. This heterogeneity often represents a severe problem for traditional mapping or ground survey techniques. The work presented here proposes to apply a very detailed CORINE land cover map as baseline information for plant community sampling and analysis in a coastal dune landscape. Location: Molise coast, Central Italy. Method: We analysed through an error matrix the coherence between land cover classes and vegetation types identified through a field survey. The CORINE land cover map (scale 1: 5000) of the Molise coast was used with the CORINE legend expanded to a fourth level of detail for natural and semi‐natural areas. Vegetation data were collected following a random stratified sampling design using the CORINE land cover classes as strata. An error matrix was used to compare, on a category‐by‐category basis, the relationship between vegetation types (obtained by cluster analyses of sampling plots) and land cover classes of the same area. Results: The coincidence between both classification approaches is quite good. Only one land cover class shows a very weak agreement with its corresponding vegetation type; this result was interpreted as being related to human disturbance. Conclusions: Since it is based on a standard land cover classification, the proposal has a potential for application to most European coastal systems. This method could represent a first step in the environmental planning of coastal systems.     

基于地形限制特征的泾河流域遥感地表覆被分类

由于在分类方法和空间分辨率等方面存在局限性,基于粗分辨率遥感数据的传统非监督分类结果在不同地物过渡带内往往误差较大。该文提出了基于地形限制特征的分类方法,在非监督分类的基础上,将非监督分类结果按照像元进行细分,并运用地形限制条件对细分后的像元进行二次判别分类。结果表明,分类精度明显提高,其中,农田和居民点分类精度的提高最为明显。这一方法使得完全同质的单元可以进行属性的变更,改善了像元空间分辨率差造成的误差;而地形限制特征的引入减少了传统非监督分类的不确定性,使模糊区域的分类有了较为明确的区分特征,提高了分类的精度。     

土地覆盖变化检测方法比较——以内蒙古草原区为例

随着对地观测技术的不断发展,遥感影像分辨率逐渐提高,促进了基于遥感影像的变化检测从传统像元级的检测向面向对象的检测转变。为了探究面向对象的变化检测方法在土地覆盖变化检测中的有效性和适用性,对面向对象的变化检测方法与常规的变化检测方法进行对比评价。以内蒙古鄂尔多斯和包头地区为试验区,选取2002年及2011年的Landsat TM/ETM+影像为数据源,比较了图像代数运算、图像变换、图像空间结构特征和面向对象的多种变化检测方法,对研究区两期土地覆盖进行了变化检测研究。结果表明:面向对象的变化检测方法在总体精度、kappa系数上都有明显的优越性,总体精度均在87.42%以上,尤其以面向对象的变化矢量分析方法精度最高,达91.56%。此外,主成分差异法也有较好的检测效果,总体精度为87.83%。对总体精度较高的3种方法在不同土地覆盖变化类型中检测效果的研究表明:对于研究区几种主要土地覆盖变化类型,面向对象的变化矢量分析法均有较理想的检测效果,平均精度为85%左右,且始终优于面向对象的光谱向量相似法,以居民地及旱地相关的变化类型最为明显;主成分差异法对不同土地覆盖变化类型检测效果差异很大,对其中4种变化类型的精度甚至达到了93%以上,但对于检测草地与裸地间转化精度很低,甚至只有8.69%;在与工矿用地有关的土地覆盖变化类型中,面向对象的变化矢量分析法的精度明显高于主成分差异法,而在与居民地有关的变化类型中,主成分差异法表现出一定优势。