BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20210916T132456Z
LOCATION:
DTSTART;TZID=Europe/Stockholm:20210706T173000
DTEND;TZID=Europe/Stockholm:20210706T190000
UID:submissions.pasc-conference.org_PASC21_sess182_post127@linklings.com
SUMMARY:P12 - Stochastic Simulations with Time-Dependent Parameters to Imp
 rove UQ in Conceptual Hydrological Models
DESCRIPTION:Poster\n\nP12 - Stochastic Simulations with Time-Dependent Par
 ameters to Improve UQ in Conceptual Hydrological Models\n\nBacci, Dal Moli
 n, Fenicia, Sukys, Reichert\n\nWater resources affect human activities in 
 many different ways. Rainfall, catchment dynamics, and human interventions
  are crucial to control water quality and availability. Hence, to make pre
 cise forecasts and take informed decisions, it is important to accurately 
 model the behavior of river basins, which are complex dynamical systems wi
 th variable and uncertain features and inputs. By modeling specific hydrol
 ogical parameters of a non-linear multi-reservoir conceptual hydrological 
 scheme with stochastic time-dependent processes, we show a way to improve 
 the traditional deterministic description of a hydrological system where t
 he characterization of the different sources of uncertainty is mainly dele
 gated to a lumped error term on the model output. We analyze the advantage
 s and challenges of the proposed stochastic approach by using Bayesian inf
 erence, numerically implemented through a Particle Markov Chain Monte Carl
 o scheme in the recently developed parallel framework SPUX. Compared to th
 e deterministic case with lumped error model, we obtain larger errors in m
 odel states, larger uncertainty in prediction than in calibration, and aut
 ocorrelated model outputs. However, the additional degrees of freedom intr
 oduced with the stochastic parameters can lead to undesirable compensation
 s of model deficits or input errors, the detection of which requires cross
 -validation simulations and careful post-processing analysis.
END:VEVENT
END:VCALENDAR