Deforestation Trends and Spatial Modelling of its Drivers in the Dry Temperate Forests of Northern Pakistan–A Case Study of Chitral

Deforestation is a major environmental challenge in the mountain areas of Pakistan. The study assessed trends in the forest cover in Chitral tehsil over the last two decades using supervised land cover classification of Landsat TM satellite images from 1992, 2000, and 2009, with a maximum likelihood algorithm. In 2009, the forest cover was 10.3% of the land area of Chitral(60,000 ha). The deforestation rate increased from 0.14% per annum in 1992–2000 to 0.54% per annum in 2000–2009, with 3,759 ha forest lost over the 17 years. The spatial drivers of deforestation were investigated using a cellular automaton modelling technique to project future forest conditions. Accessibility(elevation, slope), population density, distance to settlements, and distance to administrative boundary were strongly associated with neighbourhood deforestation. A model projection showed a further loss of 23% of existing forest in Chitral tehsil by 2030, and degradation of 8%, if deforestation continues at the present rate. Arandu Union Council, with 2212 households, will lose 85% of its forest. Local communities have limited income resources and high poverty and are heavily dependent on non-timber forest products for their livelihoods. Continued deforestation will further worsen their livelihood conditions, thus improved conservation efforts are essential.     

Multi-temporal forest cover dynamics in Kashmir Himalayan region for assessing deforestation and forest degradation in the context of REDD+ policy

The role of forests is being actively considered under the agenda of REDD+ (Reducing Emissions from Deforestation and Forest Degradation plus) aimed at reducing emissions related to changes in forest cover and forest quality. Forests in general have undergone negative changes in the past in the form of deforestation and degradation, while in some countries positive changes are reported in the form of conservation, sustainable management of forests and enhancement of carbon stock. The present study in the Kashmir Himalayan forests is an effort to assess historical forest cover changes that took place from 1980 to 2009 and to predict the same for 2030 on the basis of past trend using geospatial modeling approach. Landsat data (Multispectral Scanner (MSS), Thematic Mapper (TM) and Enhanced Thematic Mapper (ETM+)) was used for the years 1980, 1990 and (2001, 2009) respectively and change detection analysis between the dates was performed. The maps generated were validated through ground truthing. The study area (3375.62 km²) from 1980-2009 has uffered deforestation and forest degradation of about 126 km² and 139.02 km² respectively which can be claimed under negative options of REDD+, while as the area that experienced no change (1514 km²) can be claimed under conservation. A small area (23.31 km2) observed as positive change can be claimed under positive options. The projected estimates of forest cover for 2030 showed increased deforestation and forest degradation on the basis of trend analysis using Cellular Automata (CA) Markov modeling. Despite the fact that country as a whole has registered a net positive change in the past few decades, but there are regions like Kashmir region of western Himalaya which have constantly undergoing deforestation as well as degradation in the past few decades.     

Land cover change and carbon stores in a tropical montane cloud forest in the Sierra Madre Oriental,Mexico

Tropical montane cloud forest is one of the ecosystems with the highest biomass worldwide, representing an important carbon store. Globally its deforestation index is –1.1%, but in Mexico it is higher than –3%. Carbon estimates are scarce globally, particularly in Mexico. The objective of this study was to simulate future land-cover scenarios for the Sierra Madre Oriental in Mexico, by analyzing past forest cover changes. Another objective was to estimate stored carbon in the two study areas. These objectives involve the generation of information that could be useful inputs to anti-deforestation public policy such as the REDD+ strategy. Remote sensing was used to measure land cover change and estimate carbon stocks. Satellite images from 2015, 2000 and 1986 were used, and Dinamica EGO freeware generatedmodels of future projections. Between 1986 and 2015, 5171 ha of forest were converted to pasture. The annual deforestation rates were –1.5% for Tlanchinol and –1.3% for the San Bartolo Tutotepec sites. Distance to roads and marginalization were highly correlated with deforestation. By 2030, an estimated 3608 ha of forest in these sites will have been converted to pasture. Stored carbon was estimated at 16.35 Mg C ha-1 for the Tlanchinol site and 12.7 Mg C ha-1 for the San Bartolo site. In the Sierra Madre Oriental deforestation due to land cover change(–1.4%) is higher than levels reported worldwide. Besides having high values of stored carbon(14.5 Mg C ha-1), these forests have high biodiversity. The models' outputs show that the deforestation process will continue if action is not taken to avoid the expansion of livestock pasturing. This can be done by paying incentives for forest conservation to the owners of the land. The results suggest that REDD+ is currently the most viable strategy for reducing deforestation rates in tropical montane cloud forests in Sierra Madre Oriental.     

More than Carbon Stocks： A Case Study of Ecosystem-based Benefits of REDD＋ in Indonesia

During the 15th Conference of the Parties(COP 15),Parties agreed that reducing emissions from deforestation and forest degradation and enhancing ’removals of greenhouse gas emission by forests’(REDD+) in developing countries through positive incentives under the United Nations Framework Convention on Climate Change(UNFCCC) was capable of dealing with global emissions.As REDD+ seeks to lower emissions by stopping deforestation and forest degradation with an international payment tier according to baseline scenarios,opportunities for ecosystem benefits such as slowing habitat fragmentation,conservation of forest biodiversity,soil conservation may be also part of this effort.The primary objective of this study is to evaluate ecosystem-based benefits of REDD+,and to identify the relationships with carbon stock changes.To achieve this goal,high resolution satellite images are combined with Normalized Difference Vegetation Index(NDVI) to identify historical deforestation in study area of Central Kalimantan,Indonesia.The carbon emissions for the period of 2000-2005 and 2005-2009 are 2.73 × 10 5 t CO 2 and 1.47 × 10 6 t CO 2 respectively,showing an increasing trend in recent years.Dring 2005-2009,number of patches(NP),patch density(PD),mean shape index distribution(SHAPE_MN) increased 30.8%,30.7% and 7.6%.Meanwhile,largest patch index(LPI),mean area(AREA_MN),area-weighted mean of shape index distribution(SHAPE_AM),neighbor distance(ENN_MN) and interspersion and juxtaposition index(IJI) decreased by 55.3%,29.7%,15.8%,53.4% and 21.5% respectively.The area regarding as positive correlation between carbon emissions and soil erosion was approximately 8.9 × 10 3 ha corresponding to 96.0% of the changing forest.These results support the view that there are strong synergies among carbon loss,forest fragmentation and soil erosion in tropical forests.Such mechanism of REDD+ is likely to present opportunities for multiple benefits that fall outside the scope of carbon stocks.     

Forest cover change assessment in conflict-affected areas of northwest Pakistan: the case of swat and shangla districts

This study examines the spatial and temporal forest cover changes in Swat and Shangla districts to understand the deforestation pattern in context of the recent security conflict in these districts. We used multi-resolution satellite images to assess the long term deforestation from 2001 to 2009 and also to identify episodic forest cutting areas appeared during the conflict period of Oct. 2007 - Oct. 2008. There are only 58 ha of deforestation identified during the conflict period while 1268 ha of gross annual deforestation were assessed during last eight years. Most of the deforestation patches persist around the administrative boundaries at sub-district levels (tehsils) which can be attributed to ambiguity in unclear jurisdiction between the forest official. The results highlight that the forest cutting appeared in Swat and Shangla during the conflict period is not as significant when compared with the long term deforestation pattern in the area. On the one side the results of the study are supportive to the picture that emerges from international studies which report high rate of deforestation in the country and on the other side it negates any relation between the security situation and the increasing deforestation in the north western Pakistan. The study concludes that deforestation assessments require verification by independent sources of data, such as satellite imagery to improve our understanding of deforestation processes.     

Remote sensing analysis of mangrove distribution and dynamics in Zhanjiang from 1991 to 2011

Mangrove forests provide valuable societal and ecological services and goods. However, they have been experiencing high annual rates of loss in many parts of the world. In order to evaluate a long-term wetland conservation strategy that compromises urban development with comprehensive wetland ecosystem management, remote sensing techniques were used to analyze the changing mangrove distribution in the Zhanjiang Mangrove Forest National Nature Reserve. Between 1991 and 2000, the mangrove area within the study region declined from 2 264.9 to 2 085.9 ha consistent with an annual decrease of 0.79%. However, there was an overall 34.3% increase in mangrove coverage from 2 085.9 to 2 801.8 ha between 2000 and 2011. Major causes of forest loss include local human pressures in the form of deforestation, conversion to agriculture, and natural forces such as erosion. The recent gain in mangrove forest cover is attributed to eff ective conservation management in the nature reserve area, including intensive mangrove plantation eff orts and increased local awareness of wetland conservation.     

基于独立成分分析和随机森林算法的城镇用地提取研究

城镇用地信息是联合国2030年可持续发展议程关注的重点之一。城市在世界范围内迅速扩张,快速准确地获取城镇用地信息对于政府决策具有重要作用。城镇土地覆盖信息非常复杂,包括人工建筑、树木、草地、水体等多种地表覆盖类型。基于传统人工测绘获取城镇用地信息费时费力并且难于及时更新。Landsat等遥感卫星数据为城镇用地信息提取提供了丰富的数据源。基于卫星遥感数据提取的城镇用地信息可以为未来城市的建设和管理提供基础的科学决策数据。基于监督分类方法和卫星遥感数据可快速地提取城镇用地信息,然而特征变量的选择对于高精度城镇用地信息提取尤为重要。为研究不同特征变量组合对于城镇用地信息提取的影响,以北京市为研究区,以2017年7月10日获取的Landsat 8 OLI影像为数据源,通过数据预处理、纹理提取、独立成分分析、主成分分析等得到4个维度的29个特征,选取了7种特征组合方案进行城镇用地提取。考虑随机森林算法性能稳定,分类精度高和可以方便进行特征重要性评价等优点,选择其作为监督分类算法以提取城镇用地信息,并进行了精度评定,以确定最优的城镇用地提取特征组合。研究发现：综合利用光谱特征和独立成分分析后的影像特征,提取城镇用地的总体精度为93.1%,Kappa系数为0.86,优于利用其他特征的提取结果;基于随机森林算法对数据进行训练后输出的各变量的归一化变量重要性与特征均值的标准差结果存在相似性,利用随机森林算法的变量重要性估计与特征均值折线图都可以进行变量重要性评价。     

Potential effects of future adoption of the REDD mechanism as a preventive measure against deforestation and forest degradation in North Korea

This study assesses potential effects of adaption to climate change in the future as a carbon related value using a baseline and credit approach, considering the implementation of the Reducing Emissions from Deforestation and forest Degradation(REDD) mechanism. Basic data were obtained for implementing the REDD mechanism in the Democratic People's Republic of Korea(DPRK) for scientific decision-making to prevent deforestation and forest degradation. The potential effects according to the implementation of the REDD mechanism in the DPRK based on forest status data(the latest) are as follows. If the deforestation rate is reduced to a level below 6% through a 20-year REDD mechanism beginning in 2011, 0.01–11.64 C-tons of carbon credit per ha could be issued for DPRK. Converted into CO?-tons per ha, this amounts to 0.03–42.68 CO?-tons, which translates to a minimum of 226,000 CO?-tons and a maximum of 289,082,000 CO?-tons overall for forests in DPRK. In terms of carbon price, this measures up to 1.10 million USD–1.4 billion USD, considering that the REDD carbon price in voluntary carbon markets in 2010 was around 5 USD.     

Temporal variations in carbon stock of Pinus roxburghii Sargent forests of Himachal Pradesh,India

The present study was conducted in Solan Forest Division of Himachal Pradesh covering an area of about 57,158 ha. The aim was to estimate and assess the temporal change in carbon stock of the Chil Working Circle, in two forest ranges of the Division, Solan and Dharampur, over the period of 1956-2011. The inventory data of the working plans of Solan Forest Division from 1956-1957, 1984-1985 and 2002- 2003 were used in the present study while field data for biomass estimation was collected for the year 2011. The results showed a declining trend in carbon stock over 1956-1984 period, however, an increasing trend over 1984-2002 was observed, which showed a further increase for the period 2002-2011. These fluctuating trends in the forest carbon stock can be related to increasing anthropogenic pressure on forests and the subsequent introduction of a ban on green felling envisaging efficient forest management, both of which affect the forest carbon pool significantly.     

加拿大北方森林火烧迹地遥感分析

森林火灾是加拿大北方森林地区最主要的扰动因素,对北方生态系统起着主导作用。基于MODIS数据,采用全球扰动指数算法(MGDI),对加拿大萨斯喀彻温省和亚伯达省2004-2011年的森林火烧迹地进行检测和分析。通过与MODIS标准火烧迹地产品以及加拿大林业局数据进行比较,扰动指数算法检测的火烧迹地面积比MODIS标准产品更接近于林业局的统计数据。分析表明,在2004-2011年间,由于火灾原因,整个研究区森林面积平均每年减少76.36万hm2,占该区域森林总面积的3.36%。萨斯喀彻温省平均每年燃烧的森林面积为46.83万hm2,亚伯达省为29.53万hm2。其中,2006、2008、2010和2011年是火灾的高峰年份。火烧迹地主要集中在生态交错带的北方保护区、针叶林保护区、针叶林平原区,以及北方平原东北部的伍德布法罗国家森林保护区。     

基于FLUS模型和SD模型耦合的中巴经济走廊土地利用变化多情景模拟

中巴经济走廊的规划和建设离不开对走廊沿线土地资源、生态环境空间格局及变化过程的科学认识。未来土地利用变化模拟研究,可为区域土地资源管理、生态环境可持续性和潜在风险评估等研究提供可靠的预测数据。本文通过耦合系统动力学模型（SD）和未来用地模拟模型（FLUS）,并结合中巴经济走廊建设和区域的生态环境政策等设置多种情景对中巴经济走廊进行土地利用模拟,充分发挥2个模型在宏观土地需求模拟以及微观土地分配上的优势。首先根据2009—2015年的历史数据构建并验证了区域土地利用SD-FLUS模型,然后模拟了2016—2030年中巴经济走廊区域惯性发展、投资优先以及和谐发展3种不同情景下的土地利用变化。结果表明：① 历年的总量模拟相对误差均小于9.00%,2015年喀什和巴基斯坦模拟的总体精度均达到90.00%以上、Kappa系数达到0.90以上,说明SD和FLUS耦合模型能有效模拟中巴经济走廊土地利用变化格局,适用于其土地利用变化的情景模拟;② 到2030年,不同情景之间的土地利用存在明显的差异。在3种情景下建设用地均扩张,和谐发展情景扩张速度居中,该情景下喀什建设用地增加了235.17 km²,巴基斯坦增加了4942.80 km²,而扩张最快的投资优先情景下,喀什建设用地增加了265.23 km²（惯性发展情景仅增加163.71 km²）,巴基斯坦建设用地增加了5918.91 km²（惯性发展情景仅增加2861.84 km²）;巴基斯坦和谐情景下的耕地增量（4768.60 km²）不到增长最多的惯性发展情景的一半,喀什耕地在和谐发展情景增加了604.44 km²,不到投资优先情景的3/4;3种情景中只有和谐发展情景下的林地得到了有效的恢复。总体而言,和谐发展情景兼顾了社会经济发展和生态环境保护,是3种情景中最理想的情景。模型模拟结果可为中巴经济走廊的可持续性研究和生态环境评估等提供一定的数据和方法支撑。     

Driving factors for forest fire occurrence in Durango State of Mexico: A geospatial perspective

Forest fire is one of the major causes of forest loss and therefore one of the main constraints for sustainable forest management worldwide. Identifying the driving factors and understanding the contribution of each factor are essential for the management of forest fire occurrence. The objective of this study is to identify variables that are spatially related to the occurrence and incidence of the forest fire in the State of Durango, Mexico. For this purpose, data from forest fire records for a five-year period were analyzed. The spatial correlations between forest fire occurrence and intensity of land use, susceptibility of vegetation, temperature, precipitation and slope were tested by Geographically Weighted Regression (GWR) method, under an Ordinary Least Square estimator. Results show that the spatial pattern of the forest fire in the study area is closely correlated with the intensity of land use, and land use change is one of the main explanatory variables. In addition, vegetation type and precipitation are also the main driving factors. The fitting model indicates obvious link between the variables. Forest fire was found to be the consequence of a particular combination of the environmental factors, and when these factors coexist with human activities, there is high probability of forest fire occurrence. Mandatory regulation of human activities is a key strategy for forest fire prevention.     

基于随机森林算法的草原地上生物量遥感估算方法研究

草原是我国面积最大的陆地生态系统,生物量是反映生态系统质量和功能的关键指标,准确地掌握草原生物量对草原资源合理利用、生态修复、畜牧业高质量发展都具有重要的意义和作用。本研究以内蒙古锡林郭勒盟为研究区,利用高分一号遥感卫星影像,结合216个野外样本数据,采用随机森林算法（Random Forest,RF）对草原地上生物量（Aboveground Biomass,AGB）遥感估算进行了适用性分析与应用。在运用随机森林算法的过程中,进行了K-折交叉验证、多元共线性诊断、偏效应等一系列分析,完成了随机森林模型的构建,同时,将建模结果与其它模型进行了对比,最终实现了锡林郭勒盟草原AGB的反演估算。结果表明：① 随机森林算法能够较好地规避生物量建模中自变量多元共线性的问题;② 随机森林模型在草原AGB估算中较其它模型具有更好的适用性,模型精度达85%,RMSE为202.13 kg/hm²;③ 应用构建的随机森林算法估算了研究区2017年草原AGB,从结果来看,其空间分布上呈现为自东向西逐渐递减的趋势;从草地类型上看,山地草甸类AGB单产最高,温性草原类总产量最高。研究结果将对草原生态系统监测评估和草原宏观管理具有一定的参考价值。     

基于FLUS模型的大都市区土地利用优化模拟——以深圳市为例

采用深圳市2010年、2015年和2016年的土地利用现状数据,运用FLUS模型对自然发展情景、生态安全情景和生态优化情景3种情景下2030年深圳市土地利用结构和空间布局的变化进行模拟。研究结果表明:以2010年为基期模拟2015年土地利用布局的kappa指数为0.862,模拟结果较为理想。3种情景下深圳市2030年土地利用布局既有共性也存在差异。生态优化情景在禁止建设区把部分生产性用地转变为具有重要生态功能的林地,在限制建设区严格控制新增建设用地,在集中连片的控制区内限制生产活动,比自然发展情景和生态安全情景更能达到城市建设和生态保护的双重目标。     

Forest cover dynamics in Palas Valley Kohistan,Hindu Kush-Himalayan Mountains,Pakistan

Forest cover change in the mountainous region is driven by a variety of anthropogenic and natural factors.The Hindu Kush-Himalayan Mountains has experienced a considerable vegetation cover change due to intensive human activities,such as population growth,proximate causes,accessibility,unstable political situations,government policy failure and poverty.The present study seeks to find out the impact of population growth and road network expansion on forest cover of Palas valley based on remotely sensed data and employing geospatial techniques.Changes in forest cover were determined by classifying time-series satellite images of Landsat and Sentinel 2 A.The images of October 1980,2000,2010 and 2017 were classified into six land cover classes and then the impact of population growth and accessibility on forest cover was analyzed.Furthermore,forest cover and land-use change detection map was prepared using classified images of 1980 and 2017.The data were collected mainly from field visits(ground verification),census reports,Communication and Works Department,Kohistan.Satellite imageries were obtained from the United States Geological Survey’s websites and classified in ERDAS imagine 2014 and ESRI ArcGIS 10.2.1 using supervised classification-maximum likelihood algorithm.Result of this study revealed that a substantial reduction in forest cover has taken place mainly in the proximity of human settlements.On the average,during the study period,annually more than 460 hectares of forest area has been converted into other uses.     

机器学习方法在预测泉水潜在出露位置中的应用

泉水出露受到多种因素影响,在传统地质勘查手段之外,各种模型方法及影响因子预测手段,也被越来越多地应用于泉水的研究中。本文尝试利用机器学习的方法进行泉水出露位置的预测研究。根据北京市野外调查,确定了1378个测试样本点,选取了高程、坡度、坡向、地形湿度指数、径流强度指数、距河流距离、距断裂距离、岩性、归一化植被指数及土地利用类型作为影响因子,对比了2种机器学习方法（随机森林模型、分类回归树模型）和地统计方法（证据权重模型）的预测效果。研究发现：随机森林模型的预测效果最好（Area Under Curve, AUC=0.86）,分类回归树和证据权重模型效果相当（AUC分别为0.81、0.80）;随机森林模型同时揭示,岩性、距断裂距离和距河流距离这3个影响因子对泉潜在出露的影响最大。本研究表明,在强烈人类活动影响下机器学习方法仍然具有较好的泉水出露预测能力,有望为泉水保护、恢复提供新的技术方法。     

Shrinking greenery: Land use and land cover changes in Kurram Agency,Kohi Safid Mountains of north-western Pakistan

Deforestation and other Land Use and Land Cover (LULC) changes, driven by variety of physical and anthropogenic factors, have altered the mountainous environment. Mountains around the world including northern and north western belts of Pakistan are highly sensitive to deforestation and other LULC changes, which have profound impacts on various sectors of bio-physical and socio-economic systems. Assessment of LULC changes has high significance for protection, conservation and monitoring mountainous environment. The present study is an attempt to assess the landscape changes with particular reference to forest cover depletion in Kurram Agency located in the north western mountain belt of Pakistan. For detailed comparative analysis the study area has been divided into three sections, which coincide with the present administrative divisions of the Agency, i.e., Upper, Lower and Central Kurram. Temporal span of this study covers four decades. In this study, land use map of 1970 and land sat satellite imageries of 1987, 2000 and 2014 were used as spatial data sets. The images were processed and classified into six LULC classes through geospatial packages and change detection maps were prepared for each division and time period. Findings of the study reveal two trends in the four major LULC categories. Forest and rangeland have shrunk, on average, by 15% and 7.5% respectively while, bare soil and rocks outcrops have expanded by 89% and agriculture land by 7.2% in Kurram agency. The water bodies and snow cover have minor fluctuation in its land area. Major causes of shrinking greenery is attributed to high influx of Afghan refugees and high energy demand of growing population. However, with outflow of the refugees from Kurram agency the general trend in forest cover has reverted and deforestation rate has slowed down.     

Examining Forest Net Primary Productivity Dynamics and Driving Forces in Northeastern China During 1982–2010

Forest net primary productivity （NPP） is a key parameter for forest monitoring and management. In this study, monthly and annual forest NPP in the northeastern China from 1982 to 2010 were simulated by using Carnegie-Ames-Stanford Approach （CASA） model with normalized difference vegetation index （NDVI） sequences derived from Advanced Very High Resolution Radiometer （AVHRR） Global Invento y Modeling and Mapping Studies （GIMMS） and Terra Moderate Resolution Imaging Spectroradiometer （MODIS） products. To address the problem of data inconsistency between AVHRR and MODIS data, a per-pixel unary linear regres- sion model based on least ~;quares method was developed to derive the monthly NDVI sequences. Results suggest that estimated forest NPP has mean relative error of 18.97% compared to observed NPP from forest inventory. Forest NPP in the northeastern China in- creased significantly during the twenty-nine years. The results of seasonal dynamic show that more clear increasing trend of forest NPP occurred in spring and awmnn. This study also examined the relationship between forest NPP and its driving forces including the climatic and anthropogenic factors. In spring and winter, temperature played the most pivotal role in forest NPR In autumn, precipitation acted as the most importanl factor affecting forest NPP, while solar radiation played the most important role in the summer. Evaportran- spiration had a close correlation with NPP for coniferous forest, mixed coniferous broadleaved forest, and broadleaved deciduous forest. Spatially, forest NPP in the Da Hinggan Mountains was more sensitive to climatic changes than in the other ecological functional re- gions. In addition to climalie change, the degradation and improvement of forests had important effects on forest NPP. Results in this study are helpful for understanding the regional carbon sequestration and can enrich the cases for the monitoring of vegetation during long time series.     

珠江三角洲地区登革热流行风险空间模拟与预测

近年来日益严重的登革热疫情已在中国南部地区形成疫情高发区,并对中国的公共卫生安全形成了一定的威胁。登革热主要受到区域内复杂的自然环境条件以及社会经济因素的影响,而利用地理空间分析方法和模型探究登革热疫情的影响因素,并对其未来流行风险的空间分布进行模拟,是有效开展登革热预防控制工作的重要基础。本文收集了珠江三角洲地区2010-2014年的登革热病例资料和土地利用、人口密度两种社会经济要素数据,构建土地利用回归（LUR）模型以分析登革热疫情与不同空间范围内的土地利用和人口密度之间的关系,并结合SLEUTH模型获取的2030年土地利用数据以及基于人口密度预测模型获取的2030年人口密度数据,预测珠江三角洲地区2030年登革热疫情风险的空间分布。结果表明,社会经济要素对登革热疫情空间分布的影响在不同范围内存在差异,半径分别为10、7、10、2和1 km的缓冲区内的人口密度、草地、城镇用地、林地和耕地进入LUR模型并对疫情有显著的影响（相关系数分别为0.779、-0.473、0.818、-0.642和-0.403）,所构建的LUR模型效果较好（调整R²为0.796,F=390.409,P<0.01）,留一交叉检验结果显示模型的相对均方根误差为0.7046,预测值与实测值的拟合精度达到0.7101。2030年城市空间扩展的区域主要分布在深圳、东莞以及广佛的交界地区,而登革热风险预测模型表明2030年登革热疫情风险较大的区域与珠江三角洲城镇用地占比、人口分布较高的地区有高度的一致性,尤其是广佛地区。因此,LUR模型可以较好地预测登革热疫情的空间分布,从而为当地卫生部门防控登革热提供方法支持。     

Spatial assessment of forest cover and land-use changes in the Hindu-Kush mountain ranges of northern Pakistan

Anthropogenic activities and natural processes are continuously altering the mountainous environment through deforestation, forest degradation and other land-use changes. It is highly important to assess, monitor and forecast forest cover and other land-use changes for the protection and conservation of mountainous environment. The present study deals with the assessment of forest cover and other land-use changes in the mountain ranges of Dir Kohistan in northern Pakistan, using high resolution multi-temporal SPOT-5 satellite images. The SPOT-5 satellite images of years 2004, 2007, 2010 and 2013 were acquired and classified into land-cover units. In addition, forest cover and land-use change detection map was developed using the classified maps of 2004 and 2013. The classified maps were verified through random field samples and Google Earth imagery (Quick birds and SPOT-5). The results showed that during the period 2004 to 2013 the area of forest land decreased by 6.4%, however, area of range land and agriculture land have increased by 22.1% and 2.9%, respectively. Similarly, barren land increased by 1.1%, whereas, area of snow cover/glacier is significantly decreased by 21.3%. The findings from the study will be useful for forestry and landscape planning and can be utilized by the local, provincial and national forest departments; and REDD+ policy makers in Pakistan.