Towards more useful water information - methods for fine-scale spatial estimation

Research output: ThesisDoctoral ThesisCollection of Articles


The world is facing unprecedented environmental issues caused by global environmental change. Knowledge about these issues is often created with the use of global environmental models. The solutions to the issues, however, need local-scale actions, and information at the local scale can be scarce. There is, therefore, a need to bridge the gap between the coarse model outputs and the local use case. This dissertation develops and tests methodologies that help to produce fine-scale estimates – i.e., downscaling methodologies – and understanding from a coarser starting point with limited data availability, focusing on an environmental modelling related to water scarcity estimates. The four case studies included in this dissertation cover three main methods with different aims. First, understanding of processes to explain water poverty is built with the help of a computational model. I show that a combination of geographically weighted principal component analysis and a composite index is an effective way to increase understanding of the spatial variation in those processes, as opposed to a coarse, aggregated view. Its usefulness, however, is critically dependent on the knowledge of the expert. Second, I test the capabilities of the advanced areal interpolation methods dasymetric mapping/modelling and pycnophylactic interpolation in downscaling environmental model outputs (runoff generation). I show that areal interpolation is highly useful due to its ability to address spatial errors in the downscaled runoff estimates, correcting for wider spatial autocorrelation structure in the output. The ancillary data used to estimate internal spatial variation, however, needs to provide an accurate representation of the processes dominant at the finer scale. Third, hydrological routing component of distributed global hydrological models are replaced with a higher resolution alternative, producing finer scale estimates of streamflow. I show that when multiple global hydrological model outputs are used together in an ensemble, their outputs can be used effectively in a local context. The three strategies for producing fine scale estimates, are useful because they reduce the amount of resources – time, money, data, and expertise – needed to produce locally relevant information from coarse resolution datasets, provided they are used in appropriate contexts and consider the limitations I discuss. For water scarcity assessments in particular, the methods increase the potential of identifying dominant processes driving water scarcity, and increase the usefulness of existing global hydrological data products in local contexts for various research and decision-making scenarios.
Translated title of the contributionKohti hyödyllisempää vesi-informaatiota - menetelmiä tarkempaan spatiaaliseen estimointiin
Original languageEnglish
QualificationDoctor's degree
Awarding Institution
  • Aalto University
  • Virrantaus, Kirsi-Kanerva, Supervising Professor
  • Kummu, Matti, Thesis Advisor
  • Guillaume, Joseph H.A., Thesis Advisor
Print ISBNs978-952-64-0796-8
Electronic ISBNs978-952-64-0797-5
Publication statusPublished - 2022
MoE publication typeG5 Doctoral dissertation (article)


  • geoinformatics
  • environmental modelling
  • hydrology
  • downscaling
  • process understanding
  • water scarcity
  • areal interpolation
  • dasymetric modelling


Dive into the research topics of 'Towards more useful water information - methods for fine-scale spatial estimation'. Together they form a unique fingerprint.

Cite this