Trudinger

50 %
50 %
Information about Trudinger
Education

Published on January 22, 2008

Author: Terenzio

Source: authorstream.com

Model-data fusion for the coupled carbon-water system :  Model-data fusion for the coupled carbon-water system Cathy Trudinger, Michael Raupach, Peter Briggs CSIRO Marine and Atmospheric Research, Australia and Peter Rayner LSCE, France Email: cathy.trudinger@csiro.au Outline:  Outline Model-data fusion (= data assimilation + parameter estimation) Parameter estimation with the Kalman filter Australian Water Availability Project OptIC project – Optimisation Intercomparison Model-data fusion:  Model-data fusion Model: - Process representation - Subjective, incomplete - Capable of interpolation & forecast Observations: - ‘Real world’ representation - Incomplete, patchy - No forecast capability Fusion: Optimal combination (involves model-obs mismatch & strategy to minimise) Analysis: - “Best of both worlds” - Identify model weaknesses - Forecast capability - Confidence limits Choices in model-data fusion:  Choices in model-data fusion Target variables – what model quantities to vary to match observations – e.g. initial conditions, model parameters, time-varying model quantities, forcing Cost function – measure of misfit between observations and corresponding model quantities e.g. J(targets) = (H(targets) - obs)2 + (targets - priors)2 Fusion method - search strategy Batch (non-sequential) e.g. down-gradient, global search Sequential e.g. Kalman filter Approach and issues will differ to some extent between disciplines – e.g. numerical weather prediction vs terrestrial carbon cycle The Ensemble Kalman filter:  The Ensemble Kalman filter Ensemble Kalman filter (EnKF) – sequential method that uses Monte Carlo techniques; error statistics are represented using an ensemble of model states. Initial ensemble Update using measurement t0 t1 t2 Time: Two steps: Model used to predict from one time to next Update using observation Model predicts Slide6:  Augmented state vector to be estimated contains Time-dependent model variables Time-independent model parameters State vector estimate at any time is due to observations up to that time Parameter estimation with the Ensemble Kalman filter Slide7:  Our component of Australian Water Availability project: develop a Hydrological and Terrestrial Biosphere Data Assimilation System for Australia OBSERVATIONS NDVI Monthly river flows Weather: rainfall, solar radiation, temperature MODEL Soil moisture Leaf carbon Water fluxes Carbon fluxes MODEL-DATA FUSION Ensemble Kalman Filter Down-gradient method (LM) Analysis of past, present and future water and carbon budgets Maps of soil moisture, vegetation growth Process understanding Drought assessments, national water balance PRIOR INFORMATION Initial parameter estimates Soil, vegetation types Slide8:  AWAP- Dynamic Model and Observation Model State variables (x) and dynamic model Dynamic model is of general form dx/dt = F (x, u, p) All fluxes (F) are functions F (x, u, p) = F (state vector, met forcing, params) Governing equations for state vector x = (W, CL): Soil water W: Leaf carbon CL: Observations (z) and observation model NDVI = func(CL) Catchment discharge = average of FWR + FWD [- extraction - river loss] State vector in EnKF: x = [W, CL, NDVI, Dis, params] Timestep = 1 day Spatial resolution = 5x5 km Slide9:  Southern Murray Darling Basin, Australia: "unimpaired" gauged catchments Slide10:  81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 Murrumbidgee Relative Soil Moisture (0 to 1) (Forward run with priors, no assimilation) Slide11:  Predicted and observed discharge 11 unimpaired catchments in Murrumbidgee basin 25-year time series: Jan 1981 to December 2005 (Forward run with priors, no assimilation) Slide12:  Model-data synthesis approach: - State and parameter estimation with the EnKF - Assimilate NDVI and monthly catchment discharge Why Kalman filter? - Can account for model error (stochastic component) - Consistent statistics (uncertainty analysis) - Forecast capability (with uncertainty) Issues: - Time-averaged observations in EnKF (e.g. monthly catchment discharge) - Specifying statistical model (model and observation errors) - KF (sequential) vs batch parameter estimation methods? (using Levenberg-Marquardt method; also OptIC project) Slide13:  Estimated parameters Monthly mean discharge/runoff Preliminary results: Adelong Creek Blue = Ensemble Kalman filter (sequential) Red = Levenberg-Marquardt (PEST) (batch) OptIC project Optimisation method intercomparison:  OptIC project Optimisation method intercomparison International intercomparison of parameter estimation methods in biogeochemistry Simple test model, noisy pseudo-data 9 participants submitted results Methods used: Down-gradient (Levenberg-Marquardt, adjoint), Sequential (extended Kalman filter, ensemble Kalman filter) Global search (Metropolis, Metropolis MCMC, Metropolis-Hastings MCMC). Slide15:  where F(t) – forcing (log-Markovian i.e. log of forcing is Markovian) x1 – fast store x2 – slow store p1, p2 – scales for effect of x1 and x2 limitation of production k1, k2 – decay rates for pools s0 – seed production (constant value to prevent collapse) OptIC model Estimate parameters p1, p2, k1, k2 Noisy pseudo-observations:  Noisy pseudo-observations T1: Gaussian (G) T4: Gaussian but noise in x2 correlated with noise in x1 (GC) T2: Log-normal (L) T6: Gaussian with 99% of x2 data missing (GM) T3: Gaussian + temporally correlated (Markov) (GT) T5: Gaussian + drifts (GD) Estimates divided by true parameters:  Estimates divided by true parameters p1 p2 k1 k2 Cost function:  Cost function Some participants used cost functions with weights, wi(t), that depended on each noisy observation zi(t) Slide19:  Down-gradient Global-search KF wi(t) = f(zi(t)) less successful than constant weights Choice of cost function:  Choice of cost function Evans (2003) – review of parameter estimation in biogeochemical models - “it was hard to find two groups of workers who made the same choice for the form of the misfit function”, with most of the differences being in the form of the weights. Evans (2003) and the OptIC project emphasise that the choice of cost function matters, and should be made deliberately not by accident or default. (Evans 2003, J. Marine Systems) Optic project results:  Optic project results Choice of cost function had large impact on results Most troublesome noise types:- temporally correlated noise The Kalman filter did as well as the batch methods For more information on OptIc: http://www.globalcarbonproject.org/ACTIVITIES/OptIC.htm Thank you!:  Thank you!

Add a comment

Related presentations

Related pages

's Truderinger Wirtshaus

Im Münchner Osten, ganz in der Nähe der Neuen Messe München, befindet sich seit 1998 das „’s Truderinger Wirtshaus“. Bester bayerischer Tradition ...
Read more

Truderinger Waldwirtschaft

A herzliches Grias God in der Truderinger Waldwirtschaft! Wir sind eine Gaststätte für große und kleine Gäste. Mit kulinarischen Kreationen aus ...
Read more

Smart Car Trudering in München - home.mobile.de

9451984 Bearbeiten Bei der Bearbeitung Ihrer Anfrage ist ein Fehler aufgetreten. Bitte versuchen Sie es erneut oder kontaktieren Sie den mobile.de ...
Read more

Neil Trudinger – Wikipedia

Neil Sidney Trudinger (* 20. Juni 1942 in Ballarat, Australien) ist ein australischer Mathematiker, der sich mit partiellen Differentialgleichungen ...
Read more

Reservierung | Truderinger Waldwirtschaft

Wir haben geöffnet. MO - SO: 11.00 bis 24.00 Uhr. SA: 10.00 bis 24.00 Uhr. Truderinger Waldwirtschaft • Wasserburger Landstr. 272 • 81827 München
Read more

Neil Trudinger - Wikipedia, the free encyclopedia

Neil Sidney Trudinger (born 20 June 1942, Ballarat, Victoria, Australia) is an Australian mathematician, known particularly for his work in the field of ...
Read more

Neil S. Trudinger - Deutsche Digitale Bibliothek

Elliptic partial differential equations of second order Gilbarg, David. - Berlin, Heidelberg, New York : Springer, 1977; Elliptic partial differential ...
Read more

Cantami - der Chor des Truderinger Musikvereins ...

CANTAMI -- Wer? derzeit 21 junge Leuten zwischen 24 und 44 Jahren - das ist CANTAMI - der Chor des Truderinger Musikvereins!
Read more

P Trudinger - Flinders University

Trudinger, P. (2008). How lonely sits the city: reading Lamentations as city and land. In Norman C. Habel and Peter Trudinger, ed. Exploring Ecological ...
Read more

Truderinger Festwoche 2016 - Frühlingsfest in München ...

Truderinger Festwoche 2016 - Infos und Termine zur Truderinger Wiesn 2016 - Programm und Bands beim Frühlingsfest Trudering - Schausteller, Festzelt und ...
Read more