| Popis: |
Local climate trends remain mostly unknown in data-deficient mountain environments because of irregularly distributed station networks and the complexity of terrain. However, there is a necessity to monitor local changes in climate, as they have direct consequences on the livelihoods of traditional mountain communities, affecting the local hydrology, food systems, and ecosystems. To overcome existing data problems and foster the development of effective climate adaptation strategies, transdisciplinary research approaches are needed. On the example of two villages in the Pamir Mountains of Tajikistan, this study aims to investigate local climate trends by synthesising information across different knowledge systems. Whereas the Pamirs are characterised by a scarce meteorological station network, local communities can possess detailed knowledge about ecological and environmental processes occurring in their immediate surroundings. Therefore, the first objective of this study is to explore communities’ knowledge of weather and climate and to examine their perceptions of climate trends. Meanwhile, climate datasets are analysed for statistical trends and attempts are made to improve their spatial resolution. As the process of knowledge integration has rarely been applied in the climate sciences, conceptual and methodological guidelines remain absent. To address this research gap, the second objective of this study concentrates on the development of a transdisciplinary research framework. In this research, a varied spatial resolution data and methodologies from various scientific fields were acquired and analysed. The spatial climate datasets Climatic Research Unit Timeseries (CRU) 4.01 and the Tropical Rainfall Measuring Mission (TRMM) 3B43 were analysed to investigate the spatiotemporal distribution of regional trends in temperature and precipitation (1950 – 2016). High-resolution temperature time series (1979 – 2018) were obtained, using a lapse rate-based statistical downscaling approach on the European Centre for Medium-Range Weather Forecast Reanalysis Fifth Generation (ERA5) dataset. In terms of snow, temporal changes in the timing and duration of the snow period were examined (2000 – 2018) using the daily snow cover product MOD10A1 from the Moderate-Resolution Imaging Spectroradiometer (MODIS). Community observations were valorised using semi-structured interviews and shared patterns of knowledge were identified by a consensus index. With regard to the second objective of this thesis, Seasonal Rounds were critically discussed as a methodology to derive information about local weather processes and to promote effective transdisciplinary research. Results showed that instrumental climate data and community observations can provide new insights and reduce uncertainties on local climate trends, as both knowledge systems refer to different scales and variables and have their own strengths and limitations. In terms of precipitation, no local data records could be obtained for the research sites, but community members expressed high agreement on decreasing levels of rain. Regarding snow, satellite observations showed a significant decrease in the length of the snow seasons in one village, whereas community members reported a decline in the absolute amount of snow as well as a delay in the timing of snow onset. The analysis of different knowledge systems was challenged when handling contrasting observations. For instance, downscaled temperature data showed a statistically significant warming trend for summer, whereas community observations showed high consent on warming temperatures in autumn and winter. Reasons behind discordances should be identified instead of prioritising one knowledge system over another. This requires collaboration across disciplinary boundaries. To foster transdisciplinary collaboration and provide more insight into annual weather events, Seasonal Rounds seemed to be a promising method. Therefore, they present a central component in the developed transdisciplinary framework. The research framework to detect local climate trends consists of four stages, including (i) problem transformation and relationship building, (ii) generation of disciplinary knowledge, (iii) knowledge integration and validation, and (iv) assessment and impact. Transdisciplinary research approaches are integral for addressing multidimensional research questions of the 21st century. To reduce remaining data uncertainties and to increase the impact of transdisciplinary climate studies, focus should be placed on establishing collaboration and mutual understanding across disciplinary boundaries as well as on the dissolution of existing power asymmetries. Whereas the approach taken in this thesis can be used to detect climate trends in other data-scarce environments, such as the Amazon or Arctic, additional research is needed to evaluate the proposed research framework in practice and to illuminate the limits of its applicability across different climatological and developmental contexts. |
