Assessment of Drought in Grasslands: Spatio – Temporal Analyses of Soil Moisture and Extreme Climate Effects in Southwestern Mongolia
by Oyudari Vova
Date of Examination:2021-05-06
Date of issue:2021-06-28
Advisor:Prof. Dr. Martin Kappas
Referee:Prof. Dr. Martin Kappas
Referee:Prof. Dr. Daniela Sauer
Referee:Prof. Dr. Heiko Faust
Referee:Prof. Dr. Tsolmon Renchin
Referee:Dr. Daniel Wyss
Referee:Dr. Pavel Groisman
Files in this item
Name:PhDThesis_OyudariVova.pdf
Size:7.64Mb
Format:PDF
Description:Main article
Abstract
English
Soil moisture plays an essential key role in the assessment of hydrological and meteorological droughts that may affect a wide area of the natural grassland and the groundwater resource. The surface soil moisture distribution as a function of time and space is highly relevant for hydrological, ecological, and agricultural applications, especially in water-limited or drought-prone regions. However, gauging soil moisture is challenging because of its high variability. While point-scale in-situ measurements are scarce, the remote sensing tools remain the only practical means to obtain regional and global-scale soil moisture estimates. A Soil Moisture and Ocean Salinity (SMOS) is the first satellite mission ever designed to gauge the Earth’s surface soil moisture (SM) at the near-daily time scales. This work aims to evaluate the spatial and temporal patterns of SMOS soil moisture, determine the effect of the climate extremes on the vegetation growth cycle, and demonstrate the feasibility of using our drought model (GDI) the Gobi Drought Index. The GDI is based on the combination of SMOS soil moisture and several products from the MODIS satellite. We used this index for hydro-meteorological drought monitoring in Southwestern Mongolia. Firstly, we validated bias-corrected SMOS soil moisture for Mongolia by the in-situ soil moisture observations 2000 to 2015. Validation shows satisfactory results for assessing drought and water-stress conditions in the grasslands of Mongolia. The correlation analysis between SMOS and Normalized Difference Vegetation Index (NDVI) index in the various ecosystems shows a high correlation between the bias-corrected, monthly-averaged SMOS and NDVI data (R2 > 0.81). Further analysis of the SMOS and in situ SM data revealed a good match between spatial SM distribution and the rainfall events over Southwestern Mongolia. For example, during dry 2015, SM was decreased by approximately 30% across the forest-steppe and steppe areas. We also notice that both NDVI and rainfall can be used as indicators for grassland monitoring in Mongolia. The second part of this research, analyzed several dzud (specific type of climate winter disaster) events (2000, 2001, 2002, and 2010) related to drought, to comprehend the spatial and temporal variability of vegetation conditions in the Gobi region of Mongolia. We determined how these extreme climatic events affect vegetation cover and local grazing conditions using the seasonal aridity index (aAIZ), NDVI, and livestock mortality data. The NDVI is used as an indicator of vegetation activity and growth. Its spatial and temporal pattern is expected to reflect the changes in surface vegetation density and status induced by water-deficit conditions. The Gobi steppe areas showed the highest degree of vulnerability to climate, with a drastic decline of grassland in arid areas. We found that under certain dzud conditions, rapid regeneration of vegetation can occur. A thick snow layer acting as a water reservoir combined with high livestock losses can lead to an increase of the maximum August NDVI. The snowy winters can cause a 10 to 20-day early peak in NDVI and the following increase in vegetation growth. However, during a year with dry winter conditions, the vegetation growth phase begins later due to water deficiency and the entire year has a weaker vegetation growth. Generally, livestock loss and the reduction of grazing pressure was played a crucial role in vegetation recovery after extreme climatic events in Mongolia. At the last stage of our study, we develop an integrated Gobi drought index (GDI), derived from SMOS and LST, PET, and NDVI MODIS products. GDI can incorporate both, the meteorological and soil moisture drought patterns and sufficiently well represent overall drought conditions in the arid lands. Specifically, the monthly GDI and 1-month standardized precipitation index SPI showed significant correlations. Both of them are useful for drought monitoring in semi-arid lands. But, the SPI requires in situ data that are sparse, while the GDI is free from the meteorological network restriction. Consequently, we compared the GDI with other drought indices (VSWI, NDDI, NDWI, and in-situ SM). Comparison of these drought indices with the GDI allowed assessing the droughts’ behavior from different angles and quantified better their intensity. The GDI maps at fine-scale (< 1km) permit extending the applicability of our drought model to regional and local studies. These maps were generated from 2000 to 2018 across Southwestern Mongolia. Fine-scale GDI drought maps are currently limited to the whole territory for Mongolia but the algorithm is dynamic and can be transported to any region. The GDI drought index can be served as a useful tool for prevention services to detect extremely dry soil and vegetation conditions posing a risk of drought and groundwater resource depletion. It was able to detect the drought events that were underestimated by the National Drought Watch System in Mongolia. In summary, with the help of satellite, climatological, and geophysical data, the integrated GDI can be beneficial for vegetation drought stress characterization and can be a useful tool to monitor the effectiveness of pasture land restoration management practices for Mongolian livelihoods. The future application of the GDI can be extended to monitor potential impacts on water resources and agriculture in Mongolia, which have been impacted by long periods of drought.
Keywords: SMOS L2; soil moisture; Nash-Sutcliffe; NDVI; Precipitation; vegetation; Aridity Index; drought; livestock mortality; MODIS Products; In-Situ Soil Moisture (SM); a multiple regression analysis