A non-parametric hidden Markov model for climate state identification
Hidden Markov models (HMMs) can allow for the varying wet and dry cycles in the climate without the need to simulate supplementary climate variables. The fitting of a parametric HMM relies upon assumptions for the state conditional distributions. It is shown that inappropriate assumptions abou...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Copernicus Publications
2003-01-01
|
| Series: | Hydrology and Earth System Sciences |
| Online Access: | http://www.hydrol-earth-syst-sci.net/7/652/2003/hess-7-652-2003.pdf |
| _version_ | 1819210992970629120 |
|---|---|
| author | M. F. Lambert M. F. Lambert J. P. Whiting A. V. Metcalfe |
| author_facet | M. F. Lambert M. F. Lambert J. P. Whiting A. V. Metcalfe |
| author_sort | M. F. Lambert |
| collection | DOAJ |
| description | Hidden Markov models (HMMs) can allow for the varying wet and dry cycles in the climate without the need to simulate supplementary climate variables. The fitting of a parametric HMM relies upon assumptions for the state conditional distributions. It is shown that inappropriate assumptions about state conditional distributions can lead to biased estimates of state transition probabilities. An alternative non-parametric model with a hidden state structure that overcomes this problem is described. It is shown that a two-state non-parametric model produces accurate estimates of both transition probabilities and the state conditional distributions. The non-parametric model can be used directly or as a technique for identifying appropriate state conditional distributions to apply when fitting a parametric HMM. The non-parametric model is fitted to data from ten rainfall stations and four streamflow gauging stations at varying distances inland from the Pacific coast of Australia. Evidence for hydrological persistence, though not mathematical persistence, was identified in both rainfall and streamflow records, with the latter showing hidden states with longer sojourn times. Persistence appears to increase with distance from the coast.</p> <p style='line-height: 20px;'><b>Keywords: </b>Hidden Markov models, non-parametric, two-state model, climate states, persistence, probability distributions |
| first_indexed | 2024-12-23T06:20:00Z |
| format | Article |
| id | doaj.art-22857c7d71154d86bd126a58e86d1642 |
| institution | Directory Open Access Journal |
| issn | 1027-5606 1607-7938 |
| language | English |
| last_indexed | 2024-12-23T06:20:00Z |
| publishDate | 2003-01-01 |
| publisher | Copernicus Publications |
| record_format | Article |
| series | Hydrology and Earth System Sciences |
| spelling | doaj.art-22857c7d71154d86bd126a58e86d16422022-12-21T17:57:13ZengCopernicus PublicationsHydrology and Earth System Sciences1027-56061607-79382003-01-0175652667A non-parametric hidden Markov model for climate state identificationM. F. LambertM. F. LambertJ. P. WhitingA. V. MetcalfeHidden Markov models (HMMs) can allow for the varying wet and dry cycles in the climate without the need to simulate supplementary climate variables. The fitting of a parametric HMM relies upon assumptions for the state conditional distributions. It is shown that inappropriate assumptions about state conditional distributions can lead to biased estimates of state transition probabilities. An alternative non-parametric model with a hidden state structure that overcomes this problem is described. It is shown that a two-state non-parametric model produces accurate estimates of both transition probabilities and the state conditional distributions. The non-parametric model can be used directly or as a technique for identifying appropriate state conditional distributions to apply when fitting a parametric HMM. The non-parametric model is fitted to data from ten rainfall stations and four streamflow gauging stations at varying distances inland from the Pacific coast of Australia. Evidence for hydrological persistence, though not mathematical persistence, was identified in both rainfall and streamflow records, with the latter showing hidden states with longer sojourn times. Persistence appears to increase with distance from the coast.</p> <p style='line-height: 20px;'><b>Keywords: </b>Hidden Markov models, non-parametric, two-state model, climate states, persistence, probability distributionshttp://www.hydrol-earth-syst-sci.net/7/652/2003/hess-7-652-2003.pdf |
| spellingShingle | M. F. Lambert M. F. Lambert J. P. Whiting A. V. Metcalfe A non-parametric hidden Markov model for climate state identification Hydrology and Earth System Sciences |
| title | A non-parametric hidden Markov model for climate state identification |
| title_full | A non-parametric hidden Markov model for climate state identification |
| title_fullStr | A non-parametric hidden Markov model for climate state identification |
| title_full_unstemmed | A non-parametric hidden Markov model for climate state identification |
| title_short | A non-parametric hidden Markov model for climate state identification |
| title_sort | non parametric hidden markov model for climate state identification |
| url | http://www.hydrol-earth-syst-sci.net/7/652/2003/hess-7-652-2003.pdf |
| work_keys_str_mv | AT mflambert anonparametrichiddenmarkovmodelforclimatestateidentification AT mflambert anonparametrichiddenmarkovmodelforclimatestateidentification AT jpwhiting anonparametrichiddenmarkovmodelforclimatestateidentification AT avmetcalfe anonparametrichiddenmarkovmodelforclimatestateidentification AT mflambert nonparametrichiddenmarkovmodelforclimatestateidentification AT mflambert nonparametrichiddenmarkovmodelforclimatestateidentification AT jpwhiting nonparametrichiddenmarkovmodelforclimatestateidentification AT avmetcalfe nonparametrichiddenmarkovmodelforclimatestateidentification |