Treatment of systematic errors in land data assimilation systems

Date

2012-12-03

Author

Crow, Wt
Yılmaz, Mustafa Tuğrul

Metadata

Show full item record

Item Usage Stats

107
views

0
downloads

Data assimilation systems are generally designed to minimize the influence of random error on the estimation of system states. Yet, experience with land data assimilation systems has also revealed the presence of large systematic differences between model-derived and remotely-sensed estimates of land surface states. Such differences are commonly resolved prior to data assimilation through implementation of a pre-processing rescaling step whereby observations are scaled (or non-linearly transformed) to somehow "match" comparable predictions made by an assimilation model. While the rationale for removing systematic differences in means (i.e., bias) between models and observations is well-established, relatively little theoretical guidance is currently available to determine the appropriate treatment of higher-order moments during rescaling. This talk presents a simple analytical argument to define an optimal linear-rescaling strategy for observations prior to their assimilation into a land surface model. While a technique based on triple collocation theory is shown to replicate this optimal strategy, commonly-applied rescaling techniques (e.g., so called "least-squares regression" and "variance matching" approaches) are shown to represent only sub-optimal approximations to it. Since the triple collocation approach is likely infeasible in many real-world circumstances, general advice for deciding between various feasible (yet sub-optimal) rescaling approaches will be presented with an emphasis of the implications of this work for the case of directly assimilating satellite radiances. While the bulk of the analysis will deal with linear rescaling techniques, its extension to nonlinear cases will also be discussed.

URI

https://hdl.handle.net/11511/86443

Conference Name

American Geophysical Union, Fall Meeting 2012

Collections

Department of Civil Engineering, Conference / Seminar

Suggestions

OpenMETU
Core

Power Spectra of Constrained Codes with Level-Based Signaling: Overcoming Finite-Length Challenges Centers, Jessica; Tan, Xinyu; Hareedy, Ahmed; Calderbank, Robert (2021-08-01) In various practical systems, certain data patterns are prone to errors if written or transmitted. In magnetic recording and communication over transmission lines, data patterns causing consecutive transitions that are not sufficiently separated are prone to errors. In Flash memory with two levels per cell, data patterns causing high-low-high charge levels on adjacent cells are prone to errors. Constrained codes are used to eliminate error-prone patterns, and they can also achieve other goals. Recently, we ...
On numerical optimization theory of infinite kernel learning Ozogur-Akyuz, S.; Weber, Gerhard Wilhelm (2010-10-01) In Machine Learning algorithms, one of the crucial issues is the representation of the data. As the given data source become heterogeneous and the data are large-scale, multiple kernel methods help to classify "nonlinear data". Nevertheless, the finite combinations of kernels are limited up to a finite choice. In order to overcome this discrepancy, a novel method of "infinite" kernel combinations is proposed with the help of infinite and semi-infinite programming regarding all elements in kernel space. Look...
MODELLING OF KERNEL MACHINES BY INFINITE AND SEMI-INFINITE PROGRAMMING Ozogur-Akyuz, S.; Weber, Gerhard Wilhelm (2009-06-03) In Machine Learning (ML) algorithms, one of the crucial issues is the representation of the data. As the data become heterogeneous and large-scale, single kernel methods become insufficient to classify nonlinear data. The finite combinations of kernels are limited up to a finite choice. In order to overcome this discrepancy, we propose a novel method of "infinite" kernel combinations for learning problems with the help of infinite and semi-infinite programming regarding all elements in kernel space. Looking...
Analysis of variance in experimental design with nonnormal error distributions Senoglu, B; Tiku, ML (2001-01-01) We consider a two-way classification model with interaction and assume that the errors have a location-scale nonnormal distribution. From an application of the modified likelihood estimation, we obtain efficient and robust estimators of the parameters. We define F statistics for testing main effects and interaction. We analyze the Box-Cox data and show that the method developed in this paper gives accurate results besides being easy theoretically and computationally.
Integrated nonlinear regression analysis of tracer and well test data Akın, Serhat (Elsevier BV, 2003-08-01) One frequent observation from conventional pressure transient test analysis is that field data match mathematical models derived for homogeneous systems. This observation suggests that pressure data as presently interpreted may not contain details concerning certain reservoir heterogeneities. On the other hand, tracer tests may be more sensitive to heterogeneous elements present in the reservoir because of the convective nature of the flow test. In this study, a possible improvement of conventional pressure...

Citation Formats

W. Crow and M. T. Yılmaz, “Treatment of systematic errors in land data assimilation systems,” presented at the American Geophysical Union, Fall Meeting 2012, 2012, Accessed: 00, 2021. [Online]. Available: https://hdl.handle.net/11511/86443.