Climate-Triggered Drought as Causes for Different Degradation Types of Natural Forests: A Multitemporal Remote Sensing Analysis in NE Iran
Research output: Types of thesis › Doctoral thesis
Contributors
Abstract
Climate-triggered forest disturbances are increasing either by drought or by other climate extremes. Droughts can change the structure and function of forests in long-term or cause large-scale disturbances such as tree mortality, forest fires and insect outbreaks in short-term. Traditional approaches such as dendroclimatological surveys could retrieve the long-term responses of forest trees to drought conditions; however, they are restricted to individual trees or local forest stands. Therefore, multitemporal satellite-based approaches are progressing for holistic assessment of climate-induced forest responses from regional to global scales. However, little information exists on the efficiency of satellite data for analyzing the effects of droughts in different forest biomes and further studies on the analysis of approaches and large-scale disturbances of droughts are required. This research was accomplished for assessing satellite-derived physiological responses of the Caspian Hyrcanian broadleaves forests to climate-triggered droughts from regional to large scales in northeast Iran. \nThe 16-day physiological anomalies of rangelands and forests were analysed using MODIS-derived indices concerning water content deficit and greenness loss, and their variations were spatially assessed with monthly and inter-seasonal precipitation anomalies from 2000 to 2016. Specifically, dimensions of forest droughts were evaluated in relations with the dimensions of meteorological and hydrological droughts. Large-scale effects of droughts were explored in terms of tree mortality, insect outbreaks, and forest fires using field observations, multitemporal Landsat and TerraClimate data. Various approaches were evaluated to explore forest responses to climate hazards such as traditional regression models, spatial autocorrelations, spatial regression models, and panel data models.\nKey findings revealed that rangelands’ anomalies did show positive responses to monthly and inter-seasonal precipitation anomalies. However, forests’ droughts were highly associated with increases in temperatures and evapotranspiration and were slightly associated with the decreases in precipitation and surface water level. The hazard intensity of droughts has affected the water content of forests higher than their greenness properties. The stages of moderate to extreme dieback of trees were significantly associated with the hazard intensity of the deficit of forests’ water content. However, the stage of severe defoliation was only associated with the hazard intensity of forests’ greenness loss. Climate hazards significantly triggered insect outbreaks and forest fires. Although maximum temperatures, precipitation deficit, availability of soil moisture and forest fires of the previous year could significantly trigger insect outbreaks, the maximum temperatures were the only significant triggers of forest fires from 2010‒2017. In addition to climate factors, environmental and anthropogenic factors could control fire severity during a dry season.\nThe overall evaluation indicated the evidence of spatial associations between satellite-derived forest disturbances and climate hazards. Future studies are required to apply the approaches that could handle big-data, use the satellite data that have finer wavelengths for large-scale mapping of forest disturbances, and discriminate climate-induced forest disturbances from those that induced by other biotic and abiotic agents.
Details
Original language | English |
---|---|
Qualification level | Dr.-Ing. |
Awarding Institution | |
Supervisors/Advisors |
|
Defense Date (Date of certificate) | 28 May 2020 |
Publication status | Published - 2020 |
No renderer: customAssociatesEventsRenderPortal,dk.atira.pure.api.shared.model.researchoutput.Thesis
Keywords
Sustainable Development Goals
Keywords
- Forest drought, GIS, spatial distribution, determining factors