Identification of Unique Lakes Using Geographic Information Systems on the Example of the Nenets Autonomous Okrug

Abstract

The possibilities of using geographic information systems (GIS) and cluster methods for identifying anomalies to identify unique lakes in the Nenets Autonomous Okrug (NAO) are demonstrated. The following tasks have been solved: 1) determination of the morphometric characteristics of lakes using methods of remote sensing of the Earth; 2) identification of anomalous morphometric characteristics by mathematical methods; 3) expert evaluation of the lakes resulting from the analysis to confirm their unique characteristics for the purpose of subsequent research and assigning them a special status. The relevance of the work is caused by the vastness and inaccessibility of the northern regions, which leads to the need for preliminary identification of objects that are most interesting for expeditionary research. In protected areas, objects that differ in their anomalous characteristics may be of particular interest. The test region of the study was limited by the boundaries of specially protected natural areas of the NAO. The deciphering of the lakes was carried out using the Global Forest Change data set. Raster processing and extraction of areal characteristics of water bodies were carried out in the QGIS software environment. The entire data set was divided into several groups according to the genetic category of the surface, which were also analyzed when identifying anomalies. This approach makes it possible to identify anomalous objects within a particular landscape. For data processing, the IBM SPSS Modeler software application was used, where the anomaly search is based on a two-stage clustering model. The search for anomalies by cluster methods is based on the fact that if the values of an instance are removed from the center of the cluster by more than a certain amount, then the object is considered an anomaly. As a result of applying the TwoStep Cluster algorithm to the sample of morphometric parameters of lakes, 42 anomalous objects were identified. The expert assessment confirmed that the identified lakes are of interest for further research. The final set included such well-known lakes as Golodnaya Guba, Peschanka-To, Kuznetskoe-To, as well as a number of small water bodies that stand out sharply for their peculiar characteristics in comparison with most of the lakes in the study region. For sparsely populated and logistically complex northern territories, the use of such an approach is an important element of field work planning.