In Mongolia, distributional patterns of plant communities vary among regions, depending on the geographical features, microtopography, and changes in aridity from north to south. Depending on the vegetation distributions, grazing patterns are also distinctive in each region. It is important to understand the relationships between the grazing patterns and vegetation characteristics. This study aimed to clarify the vegetation characteristics in which each seasonal camp is located, focusing on Bayan- Ülgii province located in the westernmost part of Mongolia. Tolbo, Altai, Altantsögts, and Sagsai counties were selected as the target area. From late July to early August 2018, we interviewed to herder households living in each county (four herder households, one in each county). And, we visited their seasonal camps and collected information about the vegetation distributed around the seasonal camps. The herder households in Tolbo and Altai counties stayed at higher elevation territories during the summer season and at lower elevation territories during the fall and winter seasons. This pattern is the basic grazing pattern in the Altai Mountains based on the existing studies. On the other hand, in Altantsögts county, the herder household stayed at the similar elevation territories throughout the year. This pattern seemed to be caused by their importance to use the vegetation dominated by Carex duriuscula during summer season. In Sagsai county, the herder household stayed near the county center, which seemed to be caused by the importance of allocating efforts to the tourism industry. The common features among these grazing patterns were that herders used areas dominated by Carex and Kobresia species as their summer camp, that they took care to prepare hay, and their grazing patterns can be adjusted for the vegetation conditions of each year.

Arid regions cover 41% of global land area and are home to more than 2×10⁹ people. Advancing global warming increases the frequency of extreme weather events such as droughts and heat waves and reduces crop productivity in arid regions. Ensuring the stability of the food supply under population growth and climate change is an urgent issue; thus, research on breeding of drought-resistant crops is under way. Here we present a method developed to evaluate the phenology and moisture availability in wheat and oryzawheat (a wheat-rice hybrid) from spectral reflectance data and canopy surface temperature for the continuous evaluation of progress in wheat breeding for arid regions. A model using the normalized difference vegetation index (NDVI) adequately reproduced the phenological stages of initial crop growth, crop development, heading, and ripening. The difference between canopy surface temperature and air temperature decreased sharply after the heading date owing to the decreased sensible and increased latent heat fluxes. Seasonal changes of the wheat water stress index (WWSI) and the satellite-based aridity index (SbAI and SbAI^*) corresponded well to those of the NDVI. The former index was better able to track the trend of qualitative seasonal changes in moisture availability. Although there were few differences in phenology (NDVI and growth period) between wheat and oryzawheat, moisture availability was remarkably lower in oryzawheat in each growth period, especially during ripening.

The republic of Djibouti is a dry country with an annual maximum temperature and rainfall being 40 degrees Celsius and less than 150 mm, respectively. Geographical coordinates, groundwater depth, and condition of water use were evaluated at 17 sites within a small catchment in Dikhil. Half of the sites were established after the 1990s and unlisted in the database summarized by the government of Djibouti. The other half sites are listed but all of them had an error of geographical coordinates. Groundwater depth was analyzed using a simulation model (GETFLOWS) for the purpose of evaluating the potential of water resources, and the simulated water levels were in proportion to the observed water levels with the root mean square error (RMSE) being 9.4 m.