Fellow: ER1 - Dr Kasia Sawicka Host institution: Wageningen University Duration: 24 months Planned start date: May 2015
Actual start date: July 2015
Project title: Sampling design optimisation for uncertainty analysis of integrated water quality models (WP2, WP3, WP5, WP6)
Supervisor name: Dr Gerard Heuvelink & Dr Dick Brus PhD enrolment: N

Objectives:

  • Review software implementations of existing uncertainty propagation analysis techniques and develop uncertainty propagation software tools to be used across the QUICS project.
  • Apply the uncertainty propagation tools to the models used by ESR1. Support the ESR2 and ESR1 with the development of sampling design optimisation techniques.
  • Develop and test a statistical methodology for design-based validation of spatio-temporal results of water quality models.
  • Collaborate with ER2 and ER3 (TUD) and ER4 (Halcrow) on translating the research outcomes to protocols and manuals for end-users, and ESR5 (EAWAG) to provide guidance on uncertainty reduction through design of monitoring campaigns.
Tasks and methodology: Create model that combines geostatistical modelling methods (e.g. regression kriging) with numerical optimisation (e.g. simulated annealing). Create a method to allow Bayesian calibration of spatially complex dynamic models. Devise model for sample design optimisation to reduce the variance of model parameters. Apply new methods to SWAT tools developed at JLU.
Results: Software tools for propagating uncertainty across different scales (D2.1, D6.1). Sample design optimisation techniques (D3.1). Training materials on uncertainty propagation analysis in integrated water quality models (D5.3); Guidelines for use of uncertainty propagation tools; joint publications with ERs and ESRs from partner institutes on propagation of uncertainty and on optimal sampling design for minimisation of input and calibration uncertainty. Protocols and manuals for end-users on designing optimum sampling networks (D6.2).
Dissemination: ER1 will publish jointly with ERs and ESRs from partner institutes on uncertainty propagation analyses in integrated water quality modelling. Two presentation/year as MC Ambassador. Protocols and manuals for end-users.
Planned secondments: Secondment to JLU, 3 months duration, to support ESR1 with development of sampling design optimisation techniques. Secondment to EAWAG, 3 months duration, to collaborate with ESR5 on developing guidance on reduction of input uncertainty through design of monitoring campaigns.