Central Asia Webinar Series featuring Dr. Ranjeet John (University of South Dakota) presenting "Interdependent Dynamics of Food, Energy and Water in Kazakhstan and Mongolia: Connecting LULCC to Transitional Socioecological Systems." Hosted by the
Abstract
This study examines the interconnections of food, energy and water (FEW), as well as their interdependent dynamics under the rapid changes in climate and intensified land use in Kazakhstan (KaZ) and Mongolia (MG) over a 40-year period (1981-2020). We applied the concept, principles and methods of socioeconomic-ecological systems to guide our research at three hierarchical levels: local, provincial and national. Net primary production, evapotranspiration, snow cover and soil moisture were used as the key indicators for food production, radiation energy, and water balance, respectively, of the rangelands that support continued increases in economies, livestock, agriculture, and human development. Our premise is that the interconnections and interdependencies of FEW measures vary significantly between KaZ and MG, among the provinces within each country, and among herding landscapes. We are testing several hypotheses based on differences in socioeconomic, biophysical, institutional conditions. At national and provincial levels, we take a macro-ecosystem approach to examine the interdependencies of NPP, ET, livestock density, crop production and their spatiotemporal relationships. A Structural Equation Model (SEM) model was used as our primary tools to model complex data from a variety of remote sensing products and available socioeconomic databases. Provinces from each country were studied as the sampling units for SEM. At local herding landscapes, we applied conventional ecosystem methods to explore the direct impacts of herding practices on FEW measures. We seek to identify the direct connections between land use practices and FEW measures. The experiment was conducted in the Almaty province in Kazakhstan and Bulgan province in Mongolia provinces through comparing the changes in FEW measures of the two experimental herder groups. Two animals from each herding family were randomly selected for movement tracking by installing a GPS collar. Intensive field campaigns, monitoring stations, and household surveys are being organized through GPS tracking, ground sampling of the herding landscapes, and household surveys.
We identified hotspots of change and significant trends in NPP, ET and percent snow cover in the past two decades (2000-2020). Grassland degradation and vegetation stress was explained by proximity to towns and cities.30m resolution land cover maps from Landsat images for every 10 years from 1991 to 2020 covering critical major events in this region (e.g. departure from USSR in 1991, join WTO in 2001, etc.) were developed utilizing cloud computing power of Google Earth Engine. Those land cover maps cover six provinces/aimags in Kazakhstan and Mongolia and were linear downscaled to match MODIS land surface property products (e.g. ET, GPP) to examine the non-linear response of ET/GPP to land cover change. Classified land cover maps will be validated and available to public. Preliminary findings show land cover conversions occurs mostly in cropland and grassland but with different rates among provinces and decades. Changes in area of forest, grassland and cropland disproportionally contributed to changes in ET/GPP. Livestock movement data from Mongolia and Kazakhstan were collected and send back from field in August and September 2020 respectively. Preliminary plots of animal trajectory show very different herding patterns among herders between those two countries. In free ranching Mongolia, water access along a river seems to be a key driver for animal movement. The herders have semi-defined boundaries and avoid overlapping with each other in distinct clusters of movement. On the contrary, herds were constrained in pre-defined pasture in Kazakhstan. The spatial distribution of grazing intensity varies among herders. We will analyze and present more information as more data is collected.
NASA Land-Cover and Land-Use Change (LCLUC) Program, lcluc-support@umd.edu