@mastersthesis{Chakraborty2020,

title = {Multi-scale assessment of drought-induced forest dieback },

author = {Tanmoy Chakraborty},

editor = {Hanna Meyer and Torsten Prinz and Carlos Granell-Canut (supervisors) 



},

url = {http://hdl.handle.net/10362/94403},

year  = {2020},

date = {2020-01-31},

address = {Lisboa},

school = {Universidade Nova De Lisboa},

abstract = {Drought has been intensified over the years and will continue to worsen due to climate change. Existing works have focused their attention on crops rather than forests. Adverse effects are felt by all flora and fauna but the impact of the recent droughts on forest ecosystems is still unknown. Greater root depth allows them to withstand the immediate impacts of drought in contrast to crops and other vegetation. This study aims to see not only the interaction between drought and forest vitality from a multi-scale and temporal viewpoint while also to detect the impact of the recent 2018/19 drought on forest vitality based on remote sensing data. The data from the German Drought Monitor was used for the area-wide estimation of drought in Germany. Vegetative indices like NDVI collected from MODIS and Sentinel 2A were used to study the interactions between drought and forest vitality. Data for both have been acquired for the years 2000-2019. A long-standing time series data was decomposed and seasonally adjusted for better cross-correlation between the variables. The cross-correlation was verified by using breakpoints estimation by dividing the data into historically observed and test data. The coniferous-dominated black forest was used as a study area for a more in-depth analysis. Results showed that forest vitality was lowest one month after a severe drought, indicated by the highest decline in NDVI for all the forest types. This was verified using high resolution Sentinel images and the highest change does correspond to the month of January 2019. There was change in NDVI of over -0.5 for 80.63% of the entire study area. The change for each forest type was 81.74%, 54.42%, 84.14% for coniferous, broadleaved and mixed forests respectively. Two decades of NDVI and Soil Moisture Index (SMI) data along with Sentinel images for better area calculation because of higher resolution make this a highly effective approach to assess the impacts of drought on forest dieback. The methodology and data can be applied across the study area and with suitable drought indices can be used to assess the drought induced forest dieback across the globe. However, in-situ analysis with ecological considerations at the individual level could further the validity of the cross-correlations among forest types and drought. Reproducibility self-assessment (https://osf.io/j97zp/): 3, 2, 3, 1, 3 (input data, pre-processing, methods, computational environment, results).},

keywords = {environmental monitoring, Mastergeotech, Remote sensing},

pubstate = {published},

tppubtype = {mastersthesis}

}

@mastersthesis{Lüdtke2018,

title = {Evaluation of Geographic Data Mining Analyst for spatial data mining and remote sensing image analysis of Earth observation data},

author = {Daria Lüdtke},

editor = {Roberto Henriques and M.S. and Carlos Granell-Canut (supervisors) },

year  = {2018},

date = {2018-02-27},

address = {Castellón},

school = {Departamento de Lenguajes y Sistemas Informáticos},

keywords = {data mining, earth observation, Mastergeotech, Remote sensing},

pubstate = {published},

tppubtype = {mastersthesis}

}

@mastersthesis{Hausen2016,

title = {Array-database Model (SciDB) for Standardized Storing of Hyperspectral Satellite Images},

author = {Eias Hausen},

editor = {Ignacio Guerrero (supervisor) and Edzer Pebesma (co-supervisor) and Marco Painho (co-supervisor)},

year  = {2016},

date = {2016-02-26},

keywords = {Mastergeotech, Remote sensing, satellite images},

pubstate = {published},

tppubtype = {mastersthesis}

}