Sebnem Duzgun

Application of spatial and non-spatial data analysis in determination of the factors that impact municipal solid waste generation rates in Turkey

In studies focusing on the factors that impact solid waste generation habits and rates, the potential spatial dependency in solid waste generation data is not considered in relating the waste generation rates to its determinants. In this study, spatial dependency is taken into account in determination of the significant socio-economic and climatic factors that may be of importance for the municipal solid waste (MSW) generation rates in different provinces of Turkey. Simultaneous spatial autoregression (SAR) and geographically weighted regression (GWR) models are used for the spatial data analyses. Similar to ordinary least squares regression (OLSR), regression coefficients are global in SAR model. In other words, the effect of a given independent variable on a dependent variable is valid for the whole country. Unlike OLSR or SAR, GWR reveals the local impact of a given factor (or independent variable) on the waste generation rates of different provinces. Results show that provinces within closer neighborhoods have similar MSW generation rates. On the other hand, this spatial autocorrelation is not very high for the exploratory variables considered in the study. OLSR and SAR models have similar regression coefficients. GWR is useful to indicate the local determinants of MSW generation rates. GWR model can be utilized to plan waste management activities at local scale including waste minimization, collection, treatment, and disposal. At global scale, the MSW generation rates in Turkey are significantly related to unemployment rate and asphalt-paved roads ratio. Yet, significances of these variables may diminish at local scale for some provinces. At local scale, different factors may be important in affecting MSW generation rates.

Performance evaluation of building detection and digital surface model extraction algorithms: Outcomes of the PRRS 2008 Algorithm Performance Contest

This paper presents the initial results of the algorithm performance contest that was organized as part of the 5th IAPR Workshop on Pattern Recognition in Remote Sensing (PRRS 2008). The focus of the 2008 contest was automatic building detection and digital surface model (DSM) extraction. A QuickBird data set with manual ground truth was used for building detection evaluation, and a stereo Ikonos data set with a highly accurate reference DSM was used for DSM extraction evaluation. Nine submissions were received for the building detection task, and three submissions were received for the DSM extraction task. We provide an overview of the data sets, the summaries of the methods used for the submissions, the details of the evaluation criteria, and the results of the initial evaluation.

Evaluation of textural features for multispectral images

Remote sensing is a field that has wide use, leading to the fact that it has a great importance. Therefore performance of selected features plays a great role. In order to gain some perspective on useful textural features, we have brought together state-of-art textural features in recent literature, yet to be applied in remote sensing field, as well as presenting a comparison with traditional ones. Therefore we selected most commonly used textural features in remote sensing that are grey-level co-occurrence matrix (GLCM) and Gabor features. Other selected features are local binary patterns (LBP), edge orientation features extracted after applying steerable filter, and histogram of oriented gradients (HOG) features. Color histogram feature is also used and compared. Since most of these features are histogram-based, we have compared performance of bin-by-bin comparison with a histogram comparison method named as diffusion distance method. During obtaining performance of each feature, k-nearest neighbor classification method (k-NN) is applied.

An evaluation of potential sampling locations in a reservoir with emphasis on conserved spatial correlation structure.

In this study, kernel density estimation (KDE) was coupled with ordinary two-dimensional kriging (OK) to reduce the number of sampling locations in measurement and kriging of dissolved oxygen (DO) concentrations in Porsuk Dam Reservoir (PDR). Conservation of the spatial correlation structure in the DO distribution was a target. KDE was used as a tool to aid in identification of the sampling locations that would be removed from the sampling network in order to decrease the total number of samples. Accordingly, several networks were generated in which sampling locations were reduced from 65 to 10 in increments of 4 or 5 points at a time based on kernel density maps. DO variograms were constructed, and DO values in PDR were kriged. Performance of the networks in DO estimations were evaluated through various error metrics, standard error maps (SEM), and whether the spatial correlation structure was conserved or not. Results indicated that smaller number of sampling points resulted in loss of information in regard to spatial correlation structure in DO. The minimum representative sampling points for PDR was 35. Efficacy of the sampling location selection method was tested against the networks generated by experts. It was shown that the evaluation approach proposed in this study provided a better sampling network design in which the spatial correlation structure of DO was sustained for kriging.

Activity level of tectonic basins, western section of the North Anatolian Fault Zone, Turkey

We employed quantitative techniques to investigate tectonic activity levels and development stages of the Bolu, Yenicaga, Dortdivan, Cerkes, Ilgaz, and Tosya structural basins along the western portions of the main trace of the North Anatolian Fault Zone (NAFZ). Our methodology incorporates six morphometric indices: basin shape (basin elongation and compactness), hypsometric integral, mountain-front sinuosity, stream length gradient index, valley floor width-to-height ratio, and asymmetry factor, obtained from the digital elevation model of the region generated from 1/25,000-scale topographic maps. These indices are integrated within the framework of an analytical hierarchy process to provide relative activity level values of the individual basins. The new analyses indicate that the basins have contrasting tectonic activity characteristics. Judging from the applied indices, the relative increasing order of the tectonic basin activity is Dortdivan, Cerkes, Yenicaga, Ilgaz, Tosya, and Bolu. Among the basins located to the north of the NAFZ, the activity decreases eastwards, whereas to the south of this profound fault zone, it decreases towards the west.

An approach for detection of buildings and changes in buildings using orthophotos and point clouds: A case study of Van Erriş earthquake

Abstract This paper presents an image analysis of the Van Erciş earthquake, and demonstrates how efficiently the orthophoto images and point clouds from stereo matching data can be used for automatic detection of buildings and changes in buildings. The proposed method contains three basic steps. The first step is to classify the high-resolution pre and post event Red- Green-Blue (RGB) orthophoto images (orthoRGB) using Support Vector Machine (SVM) classification procedure to extract the building areas. In the second step, normalized Digital Surface Model (nDSM) band derived from point clouds and Digital Terrain Model (DTM) is integrated with the SVM classification (nDSM+orthoRGB). In the last step, building damage assessment is performed through a comparison between two independent classification results from pre-and post-event data. It was observed that using the nDSM band in the classification process as additional bands the accuracy of classification increases significantly.

Development of a Tsunami Inundation Map in Detecting Tsunami Risk in Gulf of Fethiye, Turkey

NAMIDANCE tsunami simulation and visualization tool is used to create tsunami inundation maps showing quantitative maximum tsunami flow depths in Fethiye. The risk of an extreme, but likely earthquake-generated tsunami is estimated at Fethiye Bay for 14 probabilistic earthquake scenarios. The bay is located 36°39′5″N 29°7′23″E, southwestern Turkey, which has coastline to the eastern Mediterranean Sea. The tsunami simulation and inundation assessment are performed in three stages: (1) formation of a digital elevation model of the region from the best available topography/bathymetry dataset, (2) estimation of a maximum credible tsunami scenario for the region and determination of related earthquake parameters, (3) high resolution tsunami simulation and computation of near shore and overland tsunami dynamics in the study area using tsunami simulation and visualization code NAMIDANCE, (4) determination of spatial distributions of tsunami characteristics (maximum water elevations, water velocities, flow depths) under the critical tsunami condition. The results are based on the most recent descriptions of potential tsunami sources, topographic and bathymetric databases, and tsunami numerical models. We present an innovative study concentrating on preparation of quantitative flow depths and inundation maps with a very high-resolution bathymetry/topographic dataset in the eastern Mediterranean. Inundation maps will be used to analyze the effects of possible tsunamis. The presented research is crucial to raising the awareness of government officials, the public, and other stake holders about the high probability of a tsunami event in Turkey. Moreover, the results of this study will help to plan for evacuation routes, establish safe zones, and assist in preparation for the tsunami, creating public awareness, and planning evacuation routes before the actual tsunami event happens.

Unsupervised land use - land cover classification for multispectral images

In this paper, land use/land cover classification of multispectral imagery with unsupervised approaches are presented. Primarily, a pixel based recognition algorithm is applied in three stages. At the first stage, water bodies are classified by using the NIR band histogram. At the second stage, combination of several vegetation indices are used to locate vegetation and at the third stage, by using Gabor filter man-made structures are classified and the unclassified fields are left. Followingly in order to increase the success rate, pixel based classification results are combined with meanshift segmentation results and a homogeneity test is applied for each segment. The segments that passed the homogeneity test are classified to corresponding class and for the rest, pixel based results are assigned. Compared to the similar works, this approach gives successful results.

Parameterization of Channelized Training Images: A Novel Approach for Multiple-Point Simulations of Fluvial Reservoirs

Aiming at analyzing the impact of Training Image (TI) uncertainty on simulated reservoir models; this study presents a novel approach for parameterizing channelized TIs. First, the channel structure is represented mathematically in two dimensions (2D) with a Sine function. Then, the parameters of the function (i.e. amplitude and phase) and the number of channels are modified to generate different 2D TIs. Next, the third dimension (Z-direction) slices are added to generate three dimensional TIs. Thus, a TI becomes a function of four parameters, namely, the number of channels, the number of waves in each channel which is controlled by the phase value, the amplitude value of waves, and the number of Z-direction slices. The generated TIs are then used to simulate a synthetic reservoir model utilizing a proposed MPS methodology. Analysis of the generated models shows that, the reservoir model is sensitive to the number of channels, number of waves and number of Z-direction slices in the TI used.