 |
| Titel |
Diagnosing the radiative forcing from a millennium historical model integration – evaluation of a commonly used method |
| VerfasserIn |
Torben Schmith, Peter Thejll, Eduardo Zorita |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250143529
|
| Publikation (Nr.) |
 EGU/EGU2017-7257.pdf |
|
|
|
|
|
| Zusammenfassung |
| A commonly applied method for calculating radiative forcing from model integrations is the two-step procedure where first the climate feedback parameter is found by a regression between the net radiative balance and the surface temperature, and this is then used for calculating the effective radiative forcing year by year (Gregory et al. 2004; Forster and Taylor 2006).
We investigate the performance of the method on a 1000 year long model integration where the applied forcing is approximately known. Generally the performance of the model is acceptable with a mean bias of 0.22 Wm-2 and a detrended rms value of 0.29 Wm-2. The bias has a trend -0.04 Wm-2 century-1. This trend is not present in a run with an energy balance model without a deep ocean and from this we conclude that the trend is related to improper spinup of the original model. |
|
|
|
|
|