![Hier klicken, um den Treffer aus der Auswahl zu entfernen](images/unchecked.gif) |
Titel |
Adaptivity Assessment of Regional Semi-Parametric VTEC Modeling to Different Data Distributions |
VerfasserIn |
Murat Durmaz, Mahmut Onur Karslioglu |
Konferenz |
EGU General Assembly 2014
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 16 (2014) |
Datensatznummer |
250093282
|
Publikation (Nr.) |
EGU/EGU2014-7871.pdf |
|
|
|
Zusammenfassung |
Semi-parametric modelling of Vertical Total Electron Content (VTEC) combines parametric
and non-parametric models into a single regression model for estimating the parameters and
functions from Global Positioning System (GPS) observations. The parametric part is related
to the Differential Code Biases (DCBs), which are fixed unknown parameters of the
geometry-free linear combination (or the so called ionospheric observable). On the other
hand, the non-parametric component is referred to the spatio-temporal distribution of VTEC
which is estimated by applying the method of Multivariate Adaptive Regression B-Splines
(BMARS). BMARS algorithm builds an adaptive model by using tensor product of
univariate B-splines that are derived from the data. The algorithm searches for best
fitting B-spline basis functions in a scale by scale strategy, where it starts adding
large scale B-splines to the model and adaptively decreases the scale for including
smaller scale features through a modified Gram-Schmidt ortho-normalization process.
Then, the algorithm is extended to include the receiver DCBs where the estimates
of the receiver DCBs and the spatio-temporal VTEC distribution can be obtained
together in an adaptive semi-parametric model. In this work, the adaptivity of regional
semi-parametric modelling of VTEC based on BMARS is assessed in different
ground-station and data distribution scenarios. To evaluate the level of adaptivity
the resulting DCBs and VTEC maps from different scenarios are compared not
only with each other but also with CODE distributed GIMs and DCB estimates . |
|
|
|
|
|