Browsing by Author "1 st Supervisor/Examiner: Prof. Dr. Gantuya Ganbat"

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Temporal and Spatial Trends of Forest Fires and Carbon Dynamics in Tujiin Nars
    (German-Mongolian Institute for Resources and Technology, 2026-06) Enkhjargal B.; 1 st Supervisor/Examiner: Prof. Dr. Gantuya Ganbat; 2 nd Supervisor/Examiner: Dr. Batsaikhan Vanchindorj; Advisor: Ms. Enkhjin Enkhbold
    Wildfires are one of the most important disturbances in boreal and forest steppe ecosystems, and are of great importance to biodiversity, carbon cycling and ecosystem resistance. In this study, a spatiotemporal analysis of wildfire dynamics, risks, emissions and post-fire carbon uptake in Selenge Province and the Tujiin Nars State Special Protected Area of northern Mongolia from 2001–2025 were performed. Wildfire occurrence and area burned are derived from the MCD64A1 burned area product processed by Google Earth Engine with verification data from the National Emergency Management Agency (NEMA) of Mongolia. LST, precipitation, wind speed, NDVI, slope, elevation and aspect were incorporated for fire risk and carbon calculation. During the period of this study (2001-2025) 605 fire detections and 7,809.29 km² of burned area are detected in Selenge Province with maximum in 2009 (1,566.46 km²) which shows peak fire year. Fire season is dominantly in Spring (March-May, 53% of total area burned) followed by Summer (37%). While in Tujiin Nars, fire occurs in three years of 2005, 2007 and 2025 with maximum burned area occurring in 2005 (83.633 km²) followed by 2025 (41.383 km²) and 2007 (36.684 km²). Most fires were observed in north and northeast of the protected area, which coincides with distribution of forest steppe vegetation and environmental gradients. Wildfire risk was evaluated using multi criteria Wildfire Risk Index (RI) considering combustibility index, topomorphological index, human occupation index, and MaxEnt approach modeling species distribution. For provincial scale, prediction performance of MaxEnt model is highly satisfactory; LST becomes dominant factor in May (43.22%), and wind speed in October (42.88%), while precipitation is consistently important throughout the study season. Locally in Tujiin Nars the main contributing environmental variables controlling fire ignition varied with fire years, moisture and topographic factors controlled the fire in 2005 fire season; extreme thermal environment was a direct driver in 2007 fire season; while wind becomes the direct dominant driving factor for fire in 2025 fire season, which contributes 62.81% of the model and has permutation importance of 72.41%. All of the MaxEnt models have high performance in predicting fire occurrences with AUC values ranged from 0.836 to 0.966. The calculation of CO2 emission from the fire events in Tujiin Nars was based on IPCC 2006 Guidelines, and despite that the cumulative burned area is mainly grasslands/steppes (160.37 km² of 166.39 km² in total, which accounts for 96.38% of total burned area), forest fire contribution to total CO2 emission was 88.8% (5,706.96 out of 6,426.57 tons), which was 2,000,000 kg/km² of fuel load compared with grassland (2,940 kg/km²) of fire. And, the results of ecosystem recovery and carbon sequestration over years (from 2005 to 2025 based on MODIS GPP data) suggest a great recovery potential of Tujiin Nars ecosystem, where mean C sequestration over year increases from 4.80 t C/ha to 6.26 t C/ha, and the total sequestration increased from 340,251.89 ton to 443,327.84 ton in 2025. These findings highlight the importance of fire prevention and ecosystem-based management for the carbon sequestration function and integrity of the Tujiin Nars State Special Protected Area under the context of climate change.