Is there a mistake in the article about future predictions in Biological conversation journal? Berthon et al. 2018

[j.wolk...@gmail.com]

Hello everybody, 

I just read an article from Berthon et al. (2018) about SDM and I don't understand one thing. 

The article is about modeling under future conditions and authors used for the final future scenario the average of 17 Global Circulation Models.

For me, it doesn't make any sense, so I just wanted to ask if it's ok, to use the average from the different GCM for the final model.

Thank you,

J.W.

Article:

Berthon, K., et al. "Assessment and prioritisation of plant species at risk from myrtle rust (Austropuccinia psidii) under current and future climates in Australia." Biological Conservation 218 (2018): 154-162.

Is there a mistake in the article about future predictions in Biological conversation? Berthon et al. 2018

[Adam Smith]

Hi JW,

The ensemble GCM forecast is known to be more accurate than any single GVM across historical time periods, and thus by inference, the future. So the ensemble is often used by climate scientists. However, it's the most accurate prediction *across the world*, not necessarily for any particular region. So often people (like us) who are interested in projecting to a particular region use the model(s) that does best in the region. Usually two or three GCMs are chosen to represent the extremes (eg, hottest, wettest, driest, etc). So to answer your original question, it's not completely off base to use the ensemble.

Adam

[Sam Veloz]

There is no expectation that an ensemble of GCM’s will more accurately predict the future than an individual model. One reason is that future predictions are sensitive to the emission scenario used. Since we know that none of the emissions scenarios are likely to be accurate, than we shouldn’t expect an ensemble of GCMs to somehow to more accurately predict the future. Look at plots of the IPCC ensembles and you will see that an extreme from one emission scenario sometimes overlaps with a mean from another emission scenario.

We have to be careful how we discuss projections from future GCMs. We can use future GCM’s to learn how species respond to changing climate conditions but we really can’t say much about the accuracy of these forecasts.

Hope that helps,
Sam

Sent from my iPhone

> --
> You received this message because you are subscribed to the Google Groups "Maxent" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to maxent+un...@googlegroups.com.
> To post to this group, send email to max...@googlegroups.com.
> Visit this group at https://groups.google.com/group/maxent.
> For more options, visit https://groups.google.com/d/optout.

[Husam El Alqamy]

I went through the discussion and I can not conclude on any end. Are GCM ensemble models recommended or recommended against? can anybody elaborate?

	Hossameldin ELALKAMY, MPhill., PhD. GIS Analyst  Conservation Data Center|  Victoria Ministry of Environment & Climate Change P. 250.614.7521 C. 778.896.3229|395 Waterfront Cres.  Victoria, BC., V8W 9M1
	| Profile

--

You received this message because you are subscribed to the Google Groups "Maxent" group.

To unsubscribe from this group and stop receiving emails from it, send an email to maxent+unsubscribe@googlegroups.com.

[Steve Research]

As a side point (and not wanting to derail the thread) is there a good paper discussing the regional value of different GCMs?

Regards,

Steve

Sent from my iPhone

On 12 Apr 2018, at 18:48, Husam El Alqamy <alq...@gmail.com> wrote:

I went through the discussion and I can not conclude on any end. Are GCM ensemble models recommended or recommended against? can anybody elaborate?

<image.jpeg>

Hossameldin ELALKAMY, MPhill., PhD. GIS Analyst  Conservation Data Center|  Victoria Ministry of Environment & Climate Change P. 250.614.7521 C. 778.896.3229|395 Waterfront Cres.  Victoria, BC., V8W 9M1

	| Profile

On Tue, Apr 10, 2018 at 4:56 AM, <j.wolk...@gmail.com> wrote:

Hello everybody, 

I just read an article from Berthon et al. (2018) about SDM and I don't understand one thing. 

The article is about modeling under future conditions and authors used for the final future scenario the average of 17 Global Circulation Models.

For me, it doesn't make any sense, so I just wanted to ask if it's ok, to use the average from the different GCM for the final model.

Thank you,

J.W.

Article:

Berthon, K., et al. "Assessment and prioritisation of plant species at risk from myrtle rust (Austropuccinia psidii) under current and future climates in Australia." Biological Conservation 218 (2018): 154-162.

Is there a mistake in the article about future predictions in Biological conversation? Berthon et al. 2018

--
You received this message because you are subscribed to the Google Groups "Maxent" group.
To unsubscribe from this group and stop receiving emails from it, send an email to maxent+unsubscribe@googlegroups.com.
To post to this group, send email to max...@googlegroups.com.
Visit this group at https://groups.google.com/group/maxent.
For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to the Google Groups "Maxent" group.

To unsubscribe from this group and stop receiving emails from it, send an email to maxent+un...@googlegroups.com.

[Samuel Veloz]

What is the goal of your analysis? Despite what I said earlier, there is nothing wrong with using an ensemble of GCM's necessarily. If you use one though, I would argue that you shouldn't imply that this is somehow a more accurate future prediction. Also worth noting that sometimes an ensemble mean will get you into a place in climatic space that none of the individual models predict. That is a little strange if you think about it.

In most of my work, I find that using the extremes from an ensemble of models to be more useful. This does require being able to summarize and communicate the results from multiple models which can be difficult depending on your audience and objectives. I tend to frame the discussion around the results being the range of possibilities that we might expect to encounter and looking for commonalities in response across extremely different models.

Sam

To unsubscribe from this group and stop receiving emails from it, send an email to maxent+un...@googlegroups.com.

[Steven Gonzalez]

I would contend that arbitrarily median/mean ensembling of GCMs would arbitrarily bias your predictions. Instead, it makes sense to find GCMs that work well in hindcasting previous climatological conditions, and mean ensemble them. For instance, for my thesis I followed Conde et al (2011), which identifies 4 GCMs that each hindcast previous climatologies individually that a mean ensemble of all GCMs, to perform ENMs across Mexico. Conde et al.'s (2011) findings were presented in the Fourth Communication of México for the United Nations Framework Convention on Climate Change. I would recommend looking up similar conventions for your said study area and follow their references.