Improvement in Landsat Land-Cover Change Results Using Time Series Classifications and Multi-Temporal Land-Cover Classifications and Accuracy Comparison Using a Rule-Based Logic: A Case Study in Southern Primorsky Krai, Russia

Abstract

Primorsky Krai is a unique area where a very rich and important ecosystem provides vital life to many endangered flora and fauna. This northern temperate forest is located near the Pacific Ocean providing a moderating climate for a unique blend of taiga and broadleaved tree species. Forest management of this region has changed over the past 35 years, during the timescale of this analysis, as a result of the Soviet Union break up and new timber demand from nearby China. The Primorsky Krai region is specifically valuable due to its unique mammal species, notably the Siberian Tiger, one of the only locations on earth where they are still found. It is very important to preserve this ecosystem. To analyze how this forested region has changed, time-series Landsat data were analyzed for a representative path/row (path x, row x) footprint of 185km x 185k. Image data from 1976, 1989, 1998/1999, and 2009 were classified using a hybrid classification method. Resulting land cover maps indicate represent four important times during Russian history: during the Soviet Union (1976), near the time of transition (1989), in a post-Soviet transitioning economy (1998/1999), and during a more recent time of new management practices and new global forest demands (2009). Accuracy of the automated classification was assessed using a set of pixels for each class that had been selected based visual interpretation Landsat imagery and in comparison with very high spatial resolution data from Google Earth and other ancillary data. Maps were then compared to analyze land-cover change over the three periods 1976-1989, 1989-1999, and 1999-2009. Change direction from one land cover to another were analyzed further and checked for illogical changes that might have resulted from classification error. Multiple dates were compared with one another using a combination of logic related to land use and forest succession logic and a more general process of elimination. Classification results were improved based on accuracy assessment and change detection. This technique could prove useful with time series land-cover analysis. The main human disturbance is wetland to agriculture change in the western portion of the study site. Most of the agriculture changes occur in the lowland areas near Lake Khanka where wetlands are prevalent. Selective logging and forest succession change is more common in the eastern portion containing the dense forests of the Sikhote-Alin Mountains. Selective logging for hardwood species has historically occurred in this area and possesses a potential threat to the overall preservation of habitat area. The results, showing trends in agricultural change and logging, show overall less forest disturbance than hypothesized. This is an important finding that could support past and current logging management of this area and may also have significant implications for endangered species preservation. Two protected wetland regions of the study site; Khankaiskii and Xingkaihu in China, are analyzed further for agricultural development that may have occurred before and after preservation status. Overall most agricultural expansion near Lake Khanka occurs between 1976 and 1989, although some expansion occurs between every date. The establishments of the Khankaiskii nature preserve in 1997 and the Xingkaihu nature preserve in 2007 appears to have had a positive impact on restoring wetland areas in the region surrounding Lake Khanka. Areas of recent agricultural expansion are noted and should be studied further.