Key findings from Mongolia NCAP Project
Climate change exposure and extreme events
According to a linear trend estimation of temperature changes over the period from 1940 to 2007 annual temperature has increased by 2.1°C (Figure 1). Most of the increase is taking place in winter (3.6°C). Annual precipitation has not significantly changed, but it has decreased by 7% at a national level. The direction and magnitude of change differs from place to place and a small increase in precipitation is detected in the east, southeast and the Altai Mountain region whereas the southern part of the Khangai Mountains and the central region of Mongolia show a decrease.
Extreme temperature indices show an increase in both daily maximum and minimum temperature. However, the warming trend of the minimum temperature is higher than for the maximum temperature. Analysis of precipitation extremes shows dry periods becoming more dominant.
Figure 1 Annual mean temperature anomaly time series over Mongolia, °C
Downscaled projections using PRECIS
This study has used the PRECIS (Providing Regional Climate for Impacts Studies) Regional Climate Model to downscale global model output to the regional scale (50 to 60km).
Even though it is acknowledged that multi-model approaches are essential for understanding and appreciating uncertainties in individual model results, running a Regional Climate Model is in itself very expensive and needs considerable computer capacity, time and experience. Consequently, in the context of this study it was only possible to run one Regional Climate Model and choices needed to be made on which emission scenario to use (A2) and which GCM would be used to set out the boundary conditions for the model (HadCM3). Therefore it is suggested that the results from this study are read with care and are best interpreted in conjunction with results from similar modeling efforts in the region.
Two model runs were performed: the first one to simulate the historic climate for the period 1961 to 1990 and the second one to project future climate conditions for the period 2071 to 2100 using the A2 emission scenario.
Contrary to the trends observed in historical temperature, the PRECIS modeling results for the period 2070 to 2100 show a high temperature increase of 5 to 7ºC during the summer season. Temperature increases for the other seasons are estimated at 4.5 to 5.5ºC in winter, 3.5 to 5ºC in spring and 3.5 to 4.5ºC in autumn. There are also geographical differences with the western and northwestern parts of Mongolia showing the highest increase.
According to the PRECIS model results, summer precipitation will increase by 40 to 60mm in the mountainous region of central Mongolia, and the small areas in the northeast and southeast of the country. The rest of the country shows an increase of 20 to 40mm except for the northwestern part of Mongolia where the model shows a decrease of 20 to 40mm. This is also the area where many glaciers are situated and most rivers are being charged by the permafrost. In autumn, there is a 0 to 20mm increase in the west and 20 to 40mm in the east of Mongolia. In spring, a 0 to 20mm increase is projected in central and western Mongolia and 20 to 40mm in the east. In winter, there is a 0 to 20mm decrease across the whole territory of the country
Figure 2 Simulated mean annual precipitation, mm a) Observed climate of 1961 to 1990, b) Simulated climate of 1961 to 1990, c) Projected climate of 2071 to 2100.
The analysis of extreme events shows that dzuds and droughts are expected to become more frequent, especially in the northwest of the country and dry conditions are expected to increase over the whole territory.
The growing season in Mongolia is concentrated in July and August, therefore these months formed the focus of the study. The NDVI changes over the last 24 years for July and August are depicted in Figure 3. The NDVI values don’t show a significant trend up to 1994. However, from 1994 onwards, a decreasing trend can be observed. This trend can be observed in all natural zones and, in desert steppe and desert zones, the NDVI values are even dropping below the 0.06 threshold of no vegetation (i.e. bare soil).
To further understand trends in pasture resources, the normalized value of NDVI, ZNDVI, was calculated. This also allows comparison of the trends in pasture resources between different areas and regions in the country. The slope of Z NDVI values for July and August were negative in all sums. The lowest values were detected in the aimags of Uvs, Arkhangai, Bulgan, Khuvsgul, Tuv, Selenge, Dornod and Omnogobi. This clearly indicates that serious pasture degradation is taking place in many parts of the country. It is also clear from analysis that the length of the growing season is increasing across Mongolia.
Figure 3. Annual NDVI values averaged across the different natural zones
Typically, the causes of land and pasture degradation are various but the main cause is overgrazing and there are strong indications that the growing number of goats is rapidly adding to this problem. The results from a study carried out within the NCAP project show that, every year, the number of livestock exceeds the carrying capacity of the pastureland in 135 to 180 soums.
Based on a detailed analysis of the relationship between the vegetation and climate conditions in the different natural zones, a vegetation growth index was developed. Using this formula, threshold values of the index (Qz) were identified marking the point at which a shift of vegetation or natural zone is likely to occur. Future projections of temperature and rainfall in Mongolia were then used in combination with the threshold values to map the shifting boundaries of the different natural zones.
The results show that by the year 2070 northern boundaries of desert and steppe zones will move by 350 to 450km. This means that, over the next 60 years or so, the boundary of the desert is expected to move north at an average of 6 to 7km a year. It is clear that this will potentially have dramatic effects on the availability of pasture resources in the future.
In terms of the effects of future climate change on water resources in Mongolia, several studies have been carried out over the past years (Batima, 2000; Dagvadorj, 2000; Davaa, 2005 to 2006, Natsagdorj, 2004 to 2005; Gomboluudev, 2000 to 2006; Mijiddorj, 2006). These studies have already indicated that the water resources in Mongolia are very sensitive to even small changes in temperature and precipitation.
A study carried out by Batima et al. (2000) concluded that runoff in Mongolia will increase up to 2040 and will then start to decrease again. The study further suggests that the runoff peak flow will be about a month earlier in 2040 compared to the current situation. The appearance of ice phenomena in rivers is expected to be delayed by 3 to 10 days, while in spring period the ice cover breaking process is expected to start 5 to 15 days earlier than in the current situation. Equally, spring floods due to snow and ice melting are expected to start occurring.
The latest studies on changes in evaporation, evapotranspiration, and runoff based on the A2 emission scenario show a significant increase in evapotranspiration during the 2020 to 2080 period and even the increase of runoff can not compensate such for this unbalance in the water cycle of the river basin. For instance, in Uvs lake basin in western Mongolia, the increase of river runoff is projected to be 10 times less than the expected increases in evapotranspiration. This again indicates that river basins will be much drier in the future (G.Davaa, L.Natsagdorj et al., 2006).
Even though groundwater plays an important role in the livelihoods of many herding communities, there is relatively little known about groundwater resources in Mongolia. The available data and information is scarce and certainly insufficient to support sound groundwater policies and the design of a well network for assisting rural herding communities. This study has looked at the groundwater resources on one site in the center of Uberkhangai province.
From the study, it appears that the groundwater table has significantly dropped between 1997 and 2006, probably due to changes in precipitation patterns and lack of recharge. Between 1997 and 2006, the groundwater table dropped 2m and according to this projection, the level may drop 5 to 6m by 2040. These are significant observations and further studies will be needed in order to better understand groundwater dynamics and availability.
Socio-economic indicators for climate change
Vulnerability to climate change is not only a biophysical process, but also depends on socio-economic conditions and developments. This study has, therefore, also considered socio-economic indicators in order to understand the vulnerability of rural communities to the potential impacts of climate change.
Climate conditions affecting baseline vulnerability
Herders are living under direct risk of weather and climate. Local officials and 97.6% of the herders consider climate change and environmental change a reality in their area. When asked which aspects of their environment and climate had changed most significantly they named various elements including heavy snowfall, reduction of drinking water, frequent drought and dzud events, drying up of rivers and springs, reduction in hay making yield, reduction of feeding value of pasture land, sand movement and intensification of desertification. The herders also noted a decrease in the number of forage plant species, animal fatness and bodyweight, and consequently a reduction in the production of meat and milk as well as wool, cashmere and molt hair.
The importance of climate related events for rural herding households can be easily explained by the fact that herding households depend directly on their natural environment. For instance, when looking at the food intake by herding households, one realizes that livestock constitutes the main source of food in the form of meat and milk. Climate conditions that affect the survival of the livestock will therefore directly affect the food intake of the herding households.
Similarly, an analysis of the number of livestock per household during periods of favorable climate conditions compared to periods of less favorable climate conditions clearly shows the relationship between climate and household conditions. Between 1990 and 1999, the number of households with small herds decreased for 8 consecutive years whereas the number of households with larger herds increased year by year. These developments can be partly explained by the introduction of private livestock ownership, but also by the favorable weather conditions during this period. During the dzud years of 1999 to 2001 many livestock died and immediately there was an increase in the number of households with small herds and a decrease in the number of households with larger herds.
Next. . .