1 / 22

Observation and process error

2. Readings. Ecological Detective pps 140-145Polacheck, T., R. Hilborn and A. E. Punt. 1993. Fitting surplus production models: comparing methods and measuring uncertainty. Canadian Journal of Fisheries and Aquatic Sciences 50: 2597-2607.. 3. Observation error Process error. There are two major random componentsRandom production (process error)Randomness in observation .

osric
Download Presentation

Observation and process error

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


    1. 1 Observation and process error

    2. 2 Readings

    3. 3 Observation error Process error There are two major random components Random production (process error) Randomness in observation

    4. 4 Schnute’s simple example

    5. 5 Conclusion What we think about the relation between X and Y depends on whether we think there is error in the process or the observation

    6. 6 Logistic growth model

    7. 7 Stochastic Logistic Growth Model

    8. 8 Process error only

    9. 9 The deterministic prediction

    10. 10 Process error

    11. 11 Likelihood of w’s

    12. 12 Process error Estimate r,k,q, sigma Don’t need to assume initial conditions because we assume no observation error B1=I1/q Can only make predictions (easily) when we have a continuous time series of data

    13. 13 Observation error only

    14. 14 Likelihood of v’s

    15. 15 Observation error Estimate r,k,q, sigma, B0 We do need to assume initial conditions or estimate B0 , often we assume B0 =k Can make predictions without a continuous time series of data

    16. 16

    17. 17 the process for fitting 1. Generate deterministic data 2. Fit model to these data 3. Add observation and process error and fit 4. Now fit the “real data” 5. Compare observation and process error estimates

    18. 18 Analytic formula for sd

    19. 19 Analytic formula for q

    20. 20 Estimating both observation and process error We can try to estimate both the v’s and the w’s, and ?v and ?w. But if we do that we often end up with silly answers, ?v or ?w. becoming close to zero. If we specify one of the two sigmas or their ratio, then we can obtain good estimates This approach is now standard in age-structured population dynamics models

    21. 21 Example of an observation and process error analysis

    22. 22

    23. 23 Summary Observation and process error are almost always present It is often useful to contrast the estimates with observation error to the estimates with process error It is possible to include simultaneous observation and process error

More Related