As a novel surface modeling method, the whole computing process of HASM can be divided into three parts: deriving finite difference approximations to differential equations, establishing the sampling point equations and solving the algebra equations. The well known conjugate gradient method has been used extensively to solve this symmetric positive definite system instead of Gaussian's elimination based direct methods, especially for very large systems when parallel solution environment is preferred. However, a difficulty associated with the method of conjugate gradients is that it works well on matrices that are either well conditioned or have just a few distinct eigenvalues and the coefficient matrix of the algebra equation in HASM is ill-conditioned. In this paper, we show how to preprocess a linear system so that the matrix of coefficients assumes one of these nice forms. We give two other preconditioners, i.e. incomplete Cholesky decomposition conjugate gradient method (ICCG) and symmetric successive over relaxation-preconditioned conjugate gradient method (SSORCG), so as to improve the convergence rate of HASM. Furthermore, we give adequate consideration in storage scheme of the large sparse matrix and optimize the performance of sparse matrix-vector multiplication. The cost of the computation is also considered in each iteration. We implement and test the proposed method on a Dell OptiPlex 990MT machine. Numerical tests show that ICCG has the fastest convergence rate of HASM. We also find that both ICCG and SSORCG have much faster convergence rates than others.
Slope is the most basic terrain parameter to construct geo-science models. The accuracy analysis of the computing model is one of the most important issues in the geo-science. A lot of studies focused on error analysis of the computing model over the past few decades. The source of the error and the relationship between the error and the other parameters were studied. However, few studies were aimed to put forward a new model to decrease the error, which is more important for high resolution DEMs. By analyzing the error source of the slope computing model, this paper put forward a new model called 5 Node Second-order Finite Difference (5N-2FD). The characteristic of this model is that it is able to consider the multi-distance neighbor node: firstly it builds two difference models using different grid sizes, and then combines the two models into one with different weights. It is proved that it can improve the accuracy of the results remarkably through mathematical analysis. In this paper, a mathematical surface is selected to verify the accuracy of the new model. In order to study the influence of the resolution to the accuracy of the slope, this paper generated DEMs from the mathematical surface with different resolution, one is 1m, and the other is 5m. The results show that the new model can significantly improve the accuracy of the result compared with the common models. This study enriched the method system of digital terrain analysis, and provided slope data of high accuracy for many geo-science models. In addition, besides slope, there are many terrain parameters which are calculated through finite difference, such as aspect and various kinds of curvature, and the methods of this paper could afford some useful references in improving the accuracy of such terrain parameters.
This paper aims to develop an optimal location-allocation methodology for school site selection using GIS and integer programming. According to the nearby enrollment policy, the authors propose two linear programming models (boolean and integer) with constrains of total school number and school capacity. The models are simplified by eliminating the unreasonable school-residence links for reducing the number of decision variables and therefore solving the problems efficiently. Since the constraint matrix of the boolean model is a sparse matrix with two non-zero elements per row, it can be solved optimally with very small tolerance using branch and cut algorithm. The constraint matrix of the integer model is similar to the totally unimodualr matrix and can be solved optimally. In ArcGIS 10 geoprocessing framework, the school site-selection tool is designed by integrating ArcGIS network analysis, Coin-or linear programming modeler (PuLP) and linear programming solver Cplex 12. School site selection of a county region with 1276 resident points and 50 schools is tested successfully. The related network analysis, model building, model solving and result visualization can be implemented speedily in normal personal computer with Intel Dual-Core 2.44GHz CPU and 2GB memory. Case study shows that the mathematical models and solution method introduced in this paper are efficient, easy-to-use and practical for large-scale school location-allocation problems. The authors also argue that instead of using heuristic algorithms, many large-size location-allocation problems can be solved using branch and cut algorithm optimally or optimally with very small tolerance.
Total 870 soil samples were collected from the north of Henan Province over a 27 955 km2 area. Two subgroups with 435 samples were respectively used in soil property map-making, i.e. the content of exchangeable cations (CEC) and the total nitrogen (TN). The difference of map-making results between two subgroups was calculated. The stability among Kriging method, inverse distance weight method (IDW) and polygon value represented by point value method (PRP) were compared and its' influencing factors were discussed. The results showed that: (i) RMSE(Root Mean Square Error)and R (correlation coefficient) between measured data and predicted data could not represent the stability of map-making, namely, the returning probability of the spatial pattern of soil properties. And the result was differential in precision validation when using different ways. (ii) The stability of Kriging and IDW were significantly superior to the PRP. The area with relative difference lower than 0.3 didn't achieved 20% of the total area in Kriging and IDW mapping methods, but it achieved 51.57% in PRP mapping method. The area with a high difference level was scattered in the difference map when using the former two methods, but it was centralized and showed by big polygons when using PRP. (iii) The stability of soil property map-making results was disturbed by both sample distribution and high variability of soils in local area. Sample distribution was much important in keeping stability in Kriging method than that in IDW and PRP methods. In the two latter ways high variability among data values showed much impressive effects.
The research of spatio-temporal data model is the core of temporal GIS. Though the concept of temporal GIS has been proposed for years, there isnt any practical system till now. In order to simulate and save the temporal and spatial variation as well as apply the tranditional GIS functions, in this paper we chose to extend the existing ArcGIS vector data model to support temporal and spatial analysis functions. Firstly, we analysed the type of temporal and spatial variation and divided the common temporal and spatial variation into 8 different kinds. Then based on the analysis of variation types, we constructed a kind of spatio-temporal data model that support to save, analyse and simulate spatio-temporal changes in view of the vector data model of ArcGIS by using the base state with amendments model. Because that most multi-period spatial data are saved into multi-snapshots format, in other words, they are saved separately, its unrealistic to convert them into base state with amendments format manually. So in this paper, we also design and implement an algorithm which can extract and save temporal and spatial changes into base state with amendments format automatically from multi-period data. In order to verify the feasibility and efficiency of the data model, we designed and implemented a demo system to support the query, analysis, management and dynamic visualization of spatio-temporal data. The verify result showed that this data model could be used to support spatio-tempoal change recording, analysis and management.
Carbon cycle model is an effective tool in terrestrial ecosystem carbon cycle research, through which people can study the key process mechanism of terrestrial ecosystem carbon cycle and gain more knowledge about its temporal and spatial variation. However, there exist many problems in the application of traditional carbon cycle models, such as vast amounts of data processing, complex calculations, poor interoperability, and difficult to make it more widespread, etc. Using Web Processing Service standard which is established by Open Geospatial Consortium to publish and share carbon cycle models, realize cross-platform invocation, reuse and compose carbon cycle model service, can promote the development and application of carbon cycle models. In this paper, we followed Web Processing Service standard, designed the overall architecture of the online carbon cycle model service platform, and implemented the service platform through a series of processes: service interface designing, models and spatial analysis algorithm developing, Web Processing Service encapsulating and publishing, user interface based on browser client designing and programming, and so on. The online carbon cycle model service platform provided an amount of functions, including carbon cycle data and model service publication, service management, service invocation, service composition, asynchronous interaction between users and background processes, calculation status monitoring, model result visualization, etc. With a Vegetation Photosynthesis Model chain as an application instance, we demonstrated in detail the means to develop, encapsulate, invocate and compose carbon cycle model service found on Web Processing Service standard, meanwhile displayed functions and browser interface of the online carbon cycle model service platform. This platform can be applied to the field of scientific research, decision support, etc., to promote the use of carbon cycle models, thereby make them get rapid development.
Poyang Lake is the largest freshwater lake in China. It plays an important role on ecosystem services, such as flood regulation and storage, water conservation, etc. It is important for the evaluation of its environmental protection and ecosystem services to grasp the spatial pattern and laws of land cover distribution in the Poyang Lake region. Based on HJ images and Poyang Lake land cover data in 2005, this paper obtained the Poyang Lake region land cover data in 2010 by using GIS, RS technology and quantitatively analyzed the land cover spatial distribution pattern of the Poyang Lake region in 2010. The results show that: (1) Base on HJ images and the existing land cover data, the regional land cover data can be updated dynamically. Using the HJ images in 2010 with the 2005 land cover data, the 2010 land cover data was obtained by artificial visual interpretation. Verified by the measured GPS points, the accuracy of the land cover data is 80.4%; (2) Landscape ecology, GIS and statistical ideas and methods can support quantitative analysis on land cover spatial characteristics. The indices used in the patch area, patch shape index, adjacency index of different land cover types reflect the land cover spatial characteristics of the Poyang Lake region in different aspects; (3) Landscape indices can quantitatively explain the shape, size, number and space combination of the land cover and can reflect the characteristics of the ecological environment and the status of human activities and socio-economy. The area indices has showed that the rural settlement have the highest degree of fragmentation and dispersion. Shape indices have reflected the shape of the water body tend towards the regularization owing to the conversion from lake to land and the flood control projects; And (4) the adjacency indices of land cover types have quantitatively descripted the material exchange, interaction of the adjacent land cover types and have reflected the formation mechanism and succession process of land resources. The adjacent relationship of paddy fields and evergreen broadleaf forests have reflected that the deforestation has influenced the succession of forest land and paddy field. The adjacency relationship of paddy fields and rural settlements has recommend that farmland protection and monitoring should be focused on farmland surrounding the towns and rural areas. The adjacency indices of urban construction have provided a scientific basis for the construction of the towns. Analyses have showed that Poyang Lake is an arable land-based agricultural area, which has abundant water resources and forests, therefore, we should pay high attention to intensification of rural land use, water conservation, farmland protection, and the rational development and utilization of natural resources.
Landscape indices,as quantitative description to landform characteristics,have obvious scale effect and have being used widely. Researches on the scale effect of landscape indices have important significance in understanding the relationship between pattern and scale of landforms. Based on TM images,we choose four different geomorphologic types,i.e. plain,hill,mountain and complex landform in Yuncheng City,Shanxi Province,and seven landscape indices to analyze the multi-scale effects. Firstly,we did the resample from landform images to get ten grain sizes and twelve extent sizes; secondly,according to the response curves of landscape indices of different grain and extent sizes to the four landforms,we analyzed the grain effect and extent effect of seven common landscape indices. The result showed that: patch density (PD),landscape shape index (LSI),patch cohesion index (COHESION) show a gradual downward trend with increasing of grain size with different geomorphologic types,largest patch index (LPI),perimeter-area fractal dimension (PAFRAC),landscape contagion (CONTAG) and Shannon diversity index (SHDI) have no significant changes. All landscape indices showed complicate changes with the changed spatial extent in different geomorphologic types. Landscape shape index (LSI) gradually declined with the increasing spatial extent,the other landscape indices curve changed rather complicatedly in the smaller spatial extent and showed no significant changes in the larger spatial extent. In addition,according to the different landforms,the landscape indices have obvious distinction in different grain and extent sizes,so we can distinct the complexity of different landforms by landscape indices.
On account of the current situation of Chinas increasing demand on energy sources and mineral resources, this article firstly analyzes the challenges encountered by current mineral exploration, summarizes research status of the new mineral exploration technology both at home and abroad, and then systematically introduces the basic concept, research contents and implementation significance of the "Spectral Crust". The paper points out that in the mineral exploration, through the spectral detection of the earths surface and crust surface within the scope of 1 km, the purpose of the "Spectral Crust" plan is to realize the use of hyperspectral technology on geological mapping of the Earths land surface and remote sensing detection of mineral altered information, as well as the spectrum detection of the core, establish the crust hyperspectral database set, and meet the need of mineral exploration using hyperspectral technology. The project of the "Spectral Crust" caters to the needs of national strategy, strengthens technological innovation and application mode innovation, improves the capability of mineral resources detection by spaceborne and airborne hyperspectral remote sensing, establishes the ground spectra measurement platform and spectral scanning, compiling, and analysis platform of the core, develops the integration of aerospace, aerial, surficial and belowground core based remote sensing stereo detection system. This project has great strategic significance and prospecting value, is the top design of hyperspectral remote sensing technology development, will solve the two big challenges of direct prospecting and deep prospecting and promote the development of whole countrys mineral exploration technology.
The achievement of traditional image resolution reduction is primarily through image resampling. Multi-band wavelet is a newly developed branch of wavelet analysis, and its orthogonality, compact support and smoothness are better than the binary wavelet. In this paper we firstly described the demand of quantitative reduction of image resolution and the traditional methods on image resolution reduction. Then we used multi-band wavelet to reduce image resolution quantitatively based on multi-resolution time frequency characteristics. We did the image resolution reduction experiment through the 3-band wavelet transformation and extracted the low-frequency information of the 3-band wavelet as the resulting image that achieved the image resolution reduction. Lastly, compared with the results of common image re-sampling methods, such as the nearest method, bilinear interpolation method and derivative neighborhood averaging method qualitatively and quantitatively, we found that multi-band wavelet has obvious advantages in image resolution reduction.
Global leaf area index (LAI) products such as MODIS LAI product have relatively low spatial resolution (250 m-7 km) and thus can not meet the needs of high spatial resolution remote sensing applications. It is necessary to explore the feasibility of the algorithm based on physical model for LAI retrieval using high spatial resolution remote sensing imagery. This study utilizes the algorithm based on 4-scale model to retrieve LAI in planted forest from TM imagery. The bidirectional reflectance distribution function (BRDF) and clumping representation at canopy scale are both considered in the algorithm. A validation study is conducted with in-situ measurements of LAI in planted forest from Zhangye City, Gansu Province. For comparison, the empirical model using NDVI as predicted variable is also considered for LAI estimation. The results show that better fit was found between the LAI produced by the algorithm based on 4-scale model and measured LAI (R2=0.67,RMSE=0.50) than that between LAI predicted by NDVI and measured LAI (R2=0.59,RMSE=0.67). The accuracy of the algorithm based on 4-scale model is evidently better than that of empirical model when LAI>2. Better resistance to saturation limits of vegetation index is observed for the algorithm. Moreover, the sensitivity analysis of inversed LAI to band reflectance is carried out. For red band, LAI produced from the algorithm is more influenced by the decreasing reflectance than by the increasing condition. And LAI was more sensitive to reflectance at red band (ρred) than that at near infrared band (ρnir), with uncertainty value of reflectance range from -10% to -30%. This study proved the effectiveness of the algorithm based on 4-scale model in LAI estimation from TM imagery in planted forest and will be helpful in further development of physical models for high spatial resolution LAI retrieval.
Leaf area index (LAI) defined as one half of the total green leaf area per unit ground surface area. It is an important parameter of canopy structure, because it relates to many biophysical and physiological processes of canopy, including photosynthesis, respiration, transpiration, carbon cycling, net primary productivity, precipitation interception, and energy exchange, etc. Accurate measurement of forest leaf area index by means of remote sensing has been an important task in remote sensing research. The direct method of LAI measurement is time-consuming, labor-intensive and may destroy plants. Compared to the direct method, indirect methods by means of optical methods are quicker and more efficient. These methods are all based on the Beer-Lambert law. As an important means to validate remote sensing LAI products, indirect LAI ground measurement is the basis and standard of remote sensing inversion. However, indirect ground measurement method based on Beer-Lambert law has serious underestimation problem in forest. The derivation of Beer's law was originally in uniform gas medium, when applied to discrete vegetation measurement on a pixel scale. Its applicability has not got enough attention and validation. In this paper, by theory analysis, we find that the underestimation of leaf area index comes from the spatial heterogeneity of foliage area volume density, extinction depth and leaf angle projection function G if Beer-Lambert law is applied to LAI measurements in forest. Quantitative assessment of impact on LAI measurement from non-random distribution of canopy was made. It was shown that non-random distribution of canopy may bring 20-40% measurement error of LAI. An important conclusion is that the simple correction of Beer-Lambert law has significant limitations on the in situ forest LAI measurement. This method is not a fundamental solution to this underestimation problem, and the theories and methods for LAI indirect measurement need to be changed.
Soil salinity caused by natural or human-induced processes is a severe environmental hazard all over the world. Approximately 955 M ha of soil have been affected by primary salinization worldwide, whereas secondary salinization affects approximately 77 M ha, with 58% of these in irrigated areas. This requires careful monitoring of soil salinity and mapping of the extent of salinized soils. Microwave remote sensing can significantly contribute to the detection of changes in salt-affected surface features. This paper describes a L-band resonant cavity technique at 1.43 GHz used to measure the complex dielectric constants of salt soils which contain NaCl, Na2SO4 and NaHCO3 respectively. The purpose of the new measurements is to evaluate the real components and imaginary components of dielectric constants as a function of water content, salinity and types of salts. The measurements on the dielectric properties of different salt soils show that the real components are strongly affected by soil moisture, for they increase with the increment of water content, whereas salinity has little influence on them, they tend toward constants. The imaginary components are not only dependent on water content and salinity but also on types of salts. Besides increasing with water content, the imaginary components of the three salt soils show different increasing trends with the increment of salt content. The value of imaginary component is ε″ NaCI>ε″ NaHCO3>ε″ Na2SO4 under the condition of same saline concentration. Moreover, the slopes of the three salt soils vary with the change of water content that can distinguish the three salt soils. This study proves that L-band has a good sensitivity to soil salinity, and provides an experimental basis for the salt soil dielectric constant model building and future soil salinity monitoring using microwave remote sensing.
HJ-1A/1B satellite has high spatial and temporal resolution, so it is a potential data source for extraction of the planting area of rice and other crops. This paper we took Jianli County and its adjacent area, located on the hinterland of Jianghan Plain, as the study area. According to the growth period duration of rice, we selected multi-temporal HJ-1A/1B images in critical period of rice growth including transplanting stage and heading and seeding stage. Because land surface water index (LSWI) is sensitive to the increased surface moisture, so we use LSWI to distinguish cotton plants and construction land in paddy rices transplanting period. There are large differences between normalized difference vegetation index (NDVI) of water and plant, then we choose NDVI in paddy rices heading and seeding stage to distinguish moisture content. The time variation of NDVI of forest with paddy rice and lotus root is different, and paddy rices shortwave infrared reflectance (RIRS-B2) is smaller than lotus roots, so we use difference normalized vegetation index (DNDVI) and RIRS-B2 to differentiate paddy rice from forest and lotus root. We obtained a method to extract the planting area of rice based on the HJ satellite data. Then taking SPOT images as reference, the area accuracy and sample point accuracy are not less than 90%, and Kappa is not less than 0.80. So this proposed method can extract the rice planting area exactly.
In the past 20 years, the area of winter fallow fields showed a steady increase trend because of economic structure adjustment and rural labor transfer in China. In order to make good use of winter fallow fields rationally and efficiently, the dynamic threshold method for winter fallow fields monitoring and relative information extraction was studied based on a long-time series data of the SPOT-VEGETATION NDVI, the spatial distribution of farmland and the characteristics of crop phenology. Then, the spatial and temporal distribution of winter fallow fields in middle and lower reaches of Yangtze River from 2007 to 2008 was extracted as a case study. According to this study, the dynamic threshold method based on NDVI time series data is a good way to extract area information of winter fallow fields. The results indicated that the area of the winter fallow fields in the middle and lower reaches of the Yangtze River was 205.5 million km2, occupying 45.49% of the total cultivated area. The distribution of winter fallow fields was gathered in north and northwest part of the study area including Jiangsu, Anhui, Henan, Hubei, Hunan provinces and northwest regions of Jiangxi Province, in which Hubei, Hunan and Jiangsu provinces have the most largest area of winter fallow fields, which occupied 53.41% of the total winter fallow fields area in the middle and lower reaches of Yangtze River.
The ‘Three-River Headwater’ region at Qinghai Province is the source region of Yangtze River, Yellow River and Lancangjiang River. At here, grassland is the predominate ecosystem and alpine meadow is the main grassland type. In recent 30 years, large amount of grassland in this area suffered degradation, and therefore monitoring the alpine meadow degradation is important for grassland protection in the ‘Three-River Headwater’ region. The spectral characteristics of alpine meadow at different degradation levels are valuable information for alpine meadow classification and degradation monitoring through remote sensing. In this study, we measured the ground spectral data and surveyed grass samples in three sorts of alpine meadows at five degradation levels in August 2009 at the 'Three-River Headwaters’ Region, Qinghai Province. Then we extracted and analyzed four characteristic values of the collected spectral data: the reflectance at 557675 and 760 nanometer and the red edge slopes. The result indicates that there are little differences among 557 nanometers reflectance of the three sorts of alpine meadows at five degradation levels, but there are obvious differences among 675 nanometers reflectance, 760 nanometers reflectance and the red edge slopes. We could distinguish the three sorts of alpine meadows at five degradation levels and make effective monitoring according to those differences. We also calculated the normalized differential vegetation index (NDVI) value of alpine meadow at five degradation levels based on their spectral data and related them with above ground biomass in this study. The result suggests that above ground biomass of alpine meadows at different degradation levels are well related to its ‘red edge’ slope and NDVI, with the determination coefficient of 0.93 and 0.87 respectively. Thus, these two parameters are useful for above ground biomass estimation based on remote sensing data. The ‘red edge’ slope of alpine meadows reflectance not only could distinguish the three sorts of alpine meadows at five degradation levels effectively, but also has a better relationship with above ground biomass than NDVI value. Thus, ‘red edge’ slope of alpine meadows reflectance is essential for alpine meadows classification, degradation monitoring and biomass estimation.
Land-sourced sewage outfall is one of the main marine pollution sources. Monitoring of land-sourced sewage outfalls is the main content of marine environmental monitoring. However, due to the limitation of remote sensing technology, there are very few reports about land-sourced sewage outfalls monitoring using remote sensing technology, although it is the main information dectection technology for a large area. To meet the needs of land-sourced sewage outfalls monitoring, this paper proposed a novel land-sourced sewage outfalls monitoring method using remote sensing technology by referencing to the current "land-sourced sewage outfall and adjacent marine ecological environment monitoring Guide" (HY/T086-2005) of China. Considering the ability and the economic effectiveness of different satellite data, multi scale remote sensing data was used in this method. With multi scale remote sensing data, the observed ground water quality parameters and spectral data, water quality parameters were extracted. Then Bohai Sea water quality was assessed by water quality parameters and spectral characteristics. And the key monitoring area was found out based on the result of water quality assessment. Finally, positions of land-sourced sewage outfalls were detected using high resolution remote sensing data in the key monitoring area. Validation of this method was conducted in the circum-Bohai region, China, and high positioning accuracy was acquired. Using this method, efficiency of land-sourced sewage outfalls monitoring can be improved greatly, while the cost and workload can be substantially reduced.
In this paper we studied the information extraction technology of alteration mineral assemblages. Using the standard laboratory spectral database of U.S. Geological Survey, we rebuilt the reflectivity curve of key alteration minerals in porphyry copper alteration zones with ASTER data. After comparative analysis of these key alteration minerals spectral curves, we found the characteristic of absorption valley at the spectral 2.2 μm (which is corresponding to the B6 of ASTER data) of the argillic- and phyllic-altered mineral assemblage (which include muscovite, kaolinite, montmorillonite, alunite, illite, etc. ) , and with the characteristic of high reflection peaks at B4 and B7. The characteristic of absorption valley at the spectral 2.3 μm (which is corresponding to the B8 of ASTER data) of the propylitization altered mineral assemblage (which include chlorite, epidote, calcite, etc.,), and with the characteristic of a high reflection peaks at B4, while the B9 has a higher reflectivity than the B8. Accordingly, the principal component analysis model was developed. This model can make component analysis based on ASTER B1, B4, B6 and B7 to extract the argillic- and phyllic-altered mineral assemblage; component analysis based on ASTER B1, B3, B4 and B8 to extract propylitization altered mineral assemblage. At the same time, the process was established for the alteration mineral assemblage information extraction. Taking porphyry copper region in Arequipa Province of south Peru as an example, based on comprehensive analysis of the geological background of the study area, especially its typical alteration zoning and alteration mineral assemblages in the area, we found a typical porphyry alteration zonation in Arequipa Province. So we can take advantage of the technical processes and methods, and after ASTER data preprocessing, interfering surface features removal, principal component analysis and threshold setting, to extract argillic- and phyllic-altered mineral assemblage and propylitization altered mineral assemblage. Meanwhile, together with remote sensing interpretation and regional geochemical data, we forecasted the prospecting target of porphyry copper, and delineated 4 remote sensing exploration targets. Comparing with the verified of high-resolution satellite images and the known porphyry copper deposit, the result this study is reliable.
