A methodology of characterizing status and trend of land changes in oases: A case study of Sangong River watershed, Xinjiang, China

Land change is often studied with Markov models to develop a probability transition matrix. The existing methods dependent on such matrixes cannot effectively characterize some important aspects associated with land change such as status, direction, trend and regional variations. This study presents mathematical models to quantify these elements, defining unbalanced, quasi-balanced and balanced status, one- and two-way transitions and the rising or falling trends. Using these models and remote-sensing imageries, the landscape was studied for a case area, the oasis of Sangong River in Xinjiang, Northwest China where typical arid conditions prevail. Land expansion and contraction among various land types and for the entire oasis were analyzed for the periods of 1978–1987, 1978–1998 and 1987–1998. The changes were closely related to a strong economic growth after the land-reform campaign and adoption of the market economy in China in the 1980s to early 1990s, a process not strictly Markovian that requires stationarity and randomness. Information on land-change status and trend is important for a better understanding of the underlying driving processes but also for land-use planning and decision-making.