Identification of motion parameters from spatio-temporal atmospheric temperature patterns

Fradkin, Larissa; Knox, F.B.

doi:10.1007/bf00120035

Cited by 1 publication

(2 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…If a reasonable history match is obtained, the next step is to apply the algorithm's recursive version (mentioned in the previous section, on parameter estimation) and study the behavior of parameter estimates. Recursion may be applied in both space and time [e.g., Fradkin and Knox, 1985]. From then on, it is recommended that exploratory spatial time-series analysis be carried out, not only traditional residual analysis described in all books on regression, but a more extensive procedure involving various numerical experiments with selected structures, parameter estimation algorithms, and data arrangements, all with the aim of establishing the degree of violation of assumptions 1-6 presented above.…”

Section: On Statistical Model Assessmentmentioning

confidence: 99%

“…A lot of literature on the model adjustment approach is devoted to discussion of the effects differencing schemes have on the properties of the calculated heads. The question of how they affect the properties of parameter estimates when statistical parameter estimation is carried out is addressed byFradkin and Knox ['1985]. Their main conclusion is that, even though some decisions about differencing schemes can be made a priori, which particular scheme is best in each case can only be determined by numerical experimentation…”

mentioning

confidence: 99%

See 1 more Smart Citation

Statistical identification of hydrological distributed‐parameter systems: Theory and applications

Fradkin¹,

Dokter²

1987

Water Resources Research

View full text Add to dashboard Cite

A system identification methodology for distributed-parameter model building is described and compared with other methodologies for the modeling of groundwater reservoirs. The methodology is then applied to the analysis of New Zealand's Hutt Valley-Port Nicholson groundwater reservoir data. It is established that there are too few measurement wells to allow for the identification of a model suitable for forecasting reservoir performance. Application of the methodology does, however, indicate where additional wells should be drilled, so that such a model could be identified. Certain field parameters are identified to within 100%, and this accuracy is found to be acceptable, provided all the important features of the reservoir are represented in the model. INTRODUCTIONAn important problem in groundwater hydrology is the development of mathematical models that can be used to describe the behavior of the groundwater reservoir under different operating conditions. The groundwater reservoir, being a large natural system, is badly defined: its physics is well understood in qualitative terms, but not all of its quantitative characteristics can be easily established. The major impediment in modeling is the restriction on the number of available observations and experiments. The first is sometimes restricted in time for the lack of good historical records or in space for the lack of appropriate measurement wells. The second is restricted by the analyst's inability to control natural phenomena (e.g., rainfall), and for economic and environmental reasons, as he cannot easily control the amount of withdrawn water either.In this study, a system identification approach to groundwater reservoir model building is described and applied to the analysis of New Zealand's Hutt Valley-Port Nicholson groundwater reservoir data. The identification has been conducted with the purpose of building, data permiting, a model useful in forecasting reservoir performance under a set of different operating conditions. The Hutt Valley-Port Nicholson reservoir was chosen because it had been previously analyzed by Donaldson and Campbell [1977] with the help of a conventional model adjustment approach. Another aim of this work has therefore been to establish whether any new information could be extracted from the data using the new methodology by comparing the results with those obtained before.Any system identification methodology proceeds through three stages. The first stage, model structure identification, involves establishing the form of the (often parametric) equa-XNow at 15 tions appropriate for the system description. When these equations are parametric, the second stage, model identification, reduces to parameter estimation. During the third and final stage, model assessment, one ascertains whether the model is suitable for the purpose at hand. This approach is taken because, in philosophical terms, system identification is "a logical extension of the hypotheticodeductive procedures of the scientific method to handle the special case of bad...

show abstract

Section: On Statistical Model Assessmentmentioning

confidence: 99%

mentioning

confidence: 99%