Project Description

Fellow: ESR5 - Omar Wani Host institution: Eawag: Swiss Federal Institute of Aquatic Science and Technology Duration: 36 months Planned start date: November 2014
Actual start date: November 2014
Project title: Assessing the influence of model structure deficiencies and input errors on the uncertainty of model predictions (WP1, WP3, W6)
Supervisor name: Dr Jörg Rieckermann & Prof. Dr. Max Maurer PhD enrolment: Y (Awarded by Swiss Federal Institute of Technology Zurich)

Objectives:

  • Develop enhanced numerical approach to investigate the uncertainty of predicted pollutant loads from urbanized catchments with structurally uncertain models.
  • Test the enhanced numerical approach on a comprehensive dataset of sediment loads for the Sluzew catchment in Warsaw PL.
  • Apply the new approach to another case study catchment to show generality.
Tasks and methodology: Two separate structural error processes will be introduced to account for both uncertainties from incomplete input observations and model structure deficits. To overcome difficulties with computationally expensive models, the posterior distribution will be emulated based on Gaussian processes. The posterior distribution will be sampled efficiently with adaptive algorithms that generate stochastic processes based on sequences of transition kernels. This includes numerical analyses and computational experiments, based partly on existing datasets.
Results: Journal papers on: prediction uncertainty of structurally deficient water quality models, the role of input uncertainty and efficient algorithms for posterior sampling (D6.3); With ESR2 (WU) and ESR10 UBris produce recommendations to reduce input uncertainty in future monitoring campaigns (D3.1); Computational library in the open source language R of efficient algorithms for adaptive MCMC sampling and/or posterior emulation (D1.2); Report on model structure errors and input uncertainty in projects of Aquafin (D1.3).
Dissemination: Journal papers, new algorithms in open source language R, Reports and user manual of new algorithms, Two presentation/year as MC Ambassador. Annual conference presentations.
Planned secondments: To Aquafin (3 months) to apply approach on case study catchment. To University of Bristol (3 months) to assess input uncertainty from spatial variability of rainfall.