Zhixiao Xie

Kernel Density Estimation of traffic accidents in a network space

Abstract A standard planar Kernel Density Estimation (KDE) aims to produce a smooth density surface of spatial point events over a 2-D geographic space. However, the planar KDE may not be suited for characterizing certain point events, such as traffic accidents, which usually occur inside a 1-D linear space, the roadway network. This paper presents a novel network KDE approach to estimating the density of such spatial point events. One key feature of the new approach is that the network space is represented with basic linear units of equal network length, termed lixel (linear pixel), and related network topology. The use of lixel not only facilitates the systematic selection of a set of regularly spaced locations along a network for density estimation, but also makes the practical application of the network KDE feasible by significantly improving the computation efficiency. The approach is implemented in the ESRI ArcGIS environment and tested with the year 2005 traffic accident data and a road network in the Bowling Green, Kentucky area. The test results indicate that the new network KDE is more appropriate than standard planar KDE for density estimation of traffic accidents, since the latter covers space beyond the event context (network space) and is likely to overestimate the density values. The study also investigates the impacts on density calculation from two kernel functions, lixel lengths, and search bandwidths. It is found that the kernel function is least important in structuring the density pattern over network space, whereas the lixel length critically impacts the local variation details of the spatial density pattern. The search bandwidth imposes the highest influence by controlling the smoothness of the spatial pattern, showing local effects at a narrow bandwidth and revealing “ hot spots ” at larger or global scales with a wider bandwidth. More significantly, the idea of representing a linear network by a network system of equal-length lixel s may potentially lead the way to developing a suite of other network related spatial analysis and modeling methods.

Germination responses of Caragana korshinskii Kom. to light, temperature and water stress

Caragana korshinskii Kom. is a very important shrub species for vegetation rehabilitation in northern China for its high ecological and economic values. Experiments were conducted to determine its germination responses to (i) different temperature regimes under light and/or dark conditions, (ii) different light intensities, and (iii) different water potentials combined with varied constant temperatures. Under alternating temperatures (from 5:15 to 25:35°C), final percent germinations of Caragana korshinskii were quite similar. In dark conditions, constant temperatures resulted in lower final percent germinations than alternating temperatures. At a controlled temperature regime of 10:20°C, neither final percent germinations nor germination rates showed significant differences among varied light intensities. As water potentials were reduced from 0 (distilled water) to −0.6 MPa, final percent germinations increased slightly and reached the peak at approximately −0.6 MPa, however, the increment was not significant. Beyond −0.6 MPa, further water potential reduction led to decreased final percent germinations and few seeds could germinate at −1.4 MPa. Water stress also strongly inhibited germination at very high or low temperatures. The experimental results suggested that middle May might be a suitable time for aerial seeding for this species.

Object-based Vegetation Mapping in the Kissimmee River Watershed Using HyMap Data and Machine Learning Techniques

Accurate and informative vegetation maps are in urgent demand to support the Kissimmee-Okeechobee-Everglades ecosystem restoration project in South Florida. In this study, we evaluated the applicability of fine spatial resolution hyperspectral data collected from the HyMap sensor for both community- and species-level vegetation mapping. Informative and accurate vegetation maps were produced by combining machine learning methods (Support Vector Machines (SVM) and Random Forest (RF)), object-based image analysis techniques, and Minimum Noise Fraction (MNF) data transformation. An overall accuracy of 90% was obtained in discriminating 14 vegetation communities. Classification of a large number of species is also promising. An overall accuracy of 85% was achieved in identifying 55 species using a SVM model. The results indicate that fine spatial resolution hyperspectral data classification using such automated procedure has great potential to replace the manual interpretation of aerial photos for vegetation mapping in heterogeneous wetland ecosystems.

Fusing lidar and digital aerial photography for object-based forest mapping in the Florida Everglades

The Florida Everglades has a diverse forest community which needs to be accurately mapped to support the ongoing Comprehensive Everglades Restoration Plan (CERP). In this study, we examined whether a combination of light detection and ranging (lidar) and digital aerial photography can improve the accuracy of forest mapping in the Everglades, compared with using fine spatial resolution digital aerial photographs alone. We extracted lidar elevation and intensity features from original point cloud data at the object level to avoid the errors and uncertainties in the raster-based lidar methods. These features were combined with lidar-derived topographic information, and aerial photograph derived texture measures to map 7 forest communities in a portion of the Everglades. An overall accuracy of 71% and Kappa value of 0.64 were produced. We found that low-posting-density lidar data (i.e., <4 pts/m2) can significantly increase forest classification accuracy by providing important elevation, intensity, and topography information. It is anticipated that the modern lidar remote-sensing techniques can benefit the Everglades mapping to reduce the cost in CERP.

Data fusion and classifier ensemble techniques for vegetation mapping in the coastal Everglades

This study examined the applicability of data fusion and classifier ensemble techniques for vegetation mapping in the coastal Everglades. A framework was designed to combine these two techniques. In the framework, 20-m hyperspectral imagery collected from Airborne Visible/Infrared Imaging Spectrometer was first merged with 1-m Digital Orthophoto Quarter Quads using a proposed pixel/feature-level fusion strategy. The fused data set was then classified with an ensemble approach based on two contemporary machine learning algorithms: Random Forest and Support Vector Machine. The framework was applied to classify nine vegetation types in a portion of the coastal Everglades. An object-based vegetation map was produced with an overall accuracy of 90% and Kappa value of 0.86. Per-class classification accuracy varied from 61% for identifying buttonwood forest to 100% for identifying red mangrove scrub. The result shows that the framework is promising for automated vegetation mapping in the Everglades.

A Framework for Interpolating the Population Surface at the Residential-Housing-Unit Level

Population surface information is essential for a broad array of geographical studies. Many areal interpolation methods have been developed for creating population surface data from decennial census data. This paper describes a new framework to interpolate population to the lowest spatial level, i.e., housing units, based on high-resolution geospatial imagery. The framework is fundamentally a dasymetric mapping process, comprising: (1) determination of the basic dasymetric unit; (2) extracting the basic units; and (3) allocating population counts to each unit. An example was carried out to illustrate the promises and challenges of implementing such a framework using DOQQ, LIDAR, and parcel data.

Using geographically weighted variables for image classification

In this study, geographically weighted variables calculated for two tree species, Cryptomeria japonica (Sugi) and Chamaecyparis obtusa (Hinoki), were used in addition to spectral information to classify the two species and one mixed forest class. Spectral values (digital numbers for each band) of ‘Sugi’ and ‘Hinoki’ training samples were used to predict the spectral values for the two species at other locations using the inverse distance weighting (IDW) interpolation method. Next, the similarity between each pixels spectral values and their IDW predicted values was calculated for both of the tree species. The similarity measures are considered to be geographically weighted because nearer training samples have more of an impact on their calculation. The use of geographically weighted variables resulted in an increase in overall accuracy from 82.2% to 85.9% and an increase in the kappa coefficient from 0.740 to 0.795 for a support vector machine classification.

Mapping urban land cover types using object-based multiple endmember spectral mixture analysis

Spectral mixture analysis has been frequently applied in various fields to solve the mixed pixel problem in remote sensing. So far, all the research in mixture analysis has focused on the sub-pixel analysis, i.e., selecting endmembers and conducting mixture analysis at the pixel level. Research efforts in mixture analysis at the object level are very scarce, even though the object-based image analysis (OBIA) techniques have been well developed. In this study, we examined the applicability of object-based mixture analysis in an urban environment using a Landsat Thematic Mapper image. Informative and accurate object-based fraction maps (vegetation, impervious surface, and water) were produced by combining the OBIA and multiple endmember spectral mixture analysis (MESMA) techniques. A new approach to identifying the spectral representatives of a specific class for MESMA was developed. The accuracy of the object-based fraction maps were assessed using manual interpretation results of a 1-m digital aerial photograph. Object-based mixture analysis produced a higher accuracy than traditional pixel-based mixture analysis. This work illustrates the potential of object-based mixture analysis of moderate spatial resolution imagery in mapping heterogeneous urban environments.

A Spatial Dependence Approach to Retrieving Industrial Complexes from Digital Images

This article reports preliminary results of using a spatial dependence approach to retrieving industrial building complexes from digital aerial photographs. Because image retrieval was originally developed outside geography, this paper first discusses the principle of image retrieval in the context of geographic studies, the basic types of geographic features for retrieval, and the spatially dependent nature of geographic features. Based on these discussions, the spatial dependence approach is presented for the intended retrieval. Semivariogram is used to represent the spatial dependence of geographic features and for the subsequent retrieval of these features. Results show the effectiveness of this approach and warrant further investigations.

Geographically weighted modelling of surface salinity in Florida Bay using Landsat TM data

An effective remote-sensing approach is needed for surface salinity monitoring in Florida Bay, a typical estuarine and coastal ecosystem (ECE). Yet, the non-stationary nature of surface salinity makes it difficult to model with conventional regression methods. A geographically weighted regression (GWR) approach was proposed to model surface salinity from Landsat Thematic Mapper (TM) imagery in this study. The models were constructed and validated with spatiotemporally matched field-surveyed salinity and TM imagery collected in February 1999. The GWR models reported high coefficient of determination (R 2) values and low root mean square errors (RMSEs) in validation. A 1999 model was also used to hindcast the surface salinity with TM imagery collected in December 1998 and validated with surface salinity collected at that time. The validation reported a reasonably low RMSE. It suggests a GWR approach, with field survey and remotely sensed data, may be useful in modelling and predicting the spatial variation pattern of surface salinity in Florida Bay, and could potentially serve as a less costly alternative or a supplement to field survey currently undertaken for salinity monitoring in the coastal areas of the Greater Everglades.