search

cpfaff / ease

EASE (Essential Annotation Schema for Ecology)
0 stars 2 forks source link

Parameterize control factors in methods #32

Closed cpfaff closed 8 years ago

cpfaff commented 8 years ago

@EichenbergBEF We could also take inspiration from the geobon initiative on essential biodiversity variables for our top ten list. What do you think about these.

http://www.geobon.org/Downloads/articles/2013/201303_ebv_table.pdf

EichenbergBEF commented 8 years ago

I think the quite generic structure and naming of these factors is great and we should really stick to such a generality in determining the control facotrs (top 10 list). This is quite appealing to have very general control factor classes.

However, I think this list really lacks some important aspects. I mean there is no class allowing to refer to e.g. soil measurements (e.g. pH value, granulary usw.). e.g. "Soil physicochemical parameters" (I know "physicochemical" is more than a single concept, but let's take this for now) as a Control parameter, which the comprises "soil nutrient status", "pH-value", usw. these more generic classes can then (where really necessary) be broken down into "Soil chemical parameters" (which comprises pH-vale, nuritents etc) and "Soil physical parameters (which then may comprise soil moisture, soil gtranularity).

The classes within the Control Factors then have to be described in more detail in the vocabulary, so that the user knows what has to be categorized as what.

In general I find the approach of "being more general" (e.g. Soil physicochemical parameters) in the beginning, and then , splitting up the classes where necessary (Soil physical & Soil chemical) is quite appealing. If you start from very detailed information which you then acgglomerate from detailed -> generic, you might easily get lost in the details, whereas general -> detail may save you from beig too accurate.

Am 20.01.2016 um 13:42 schrieb Claas-Thido Pfaff:

@EichenbergBEF https://github.com/EichenbergBEF We could also use the the important variables from the geobon initiative on essential biodiversity variables for out top ten list. What do you think about these.

http://www.geobon.org/Downloads/articles/2013/201303_ebv_table.pdf

— Reply to this email directly or view it on GitHub https://github.com/cpfaff/cafe/issues/32#issuecomment-173193275.

Dr. rer. nat. David Eichenberg BEF-China research consortium Data manager (BEF China Dataportal) Tel: 0049-341-9738587 Department of Systematic Botany and Functional Biodiversity University of Leipzig Room 120 Johannisalles 21 04103 Leipzig GERMANY

cpfaff commented 8 years ago

That is a tough one. We have the same problems as we had with processes. What the hack is a control factor. And that is the problem. It is not something that we just measure and that has necessarily a unit. It might be something that we controlled in our setup which could be almost everything. Maybe we need the workshop here later as you @EichenbergBEF suggested to agree on a best of list and then agree on some sensitive attributes that could be used for a classification.

For now I will try to merge what we have with what is provide by the ebv document.

cpfaff commented 8 years ago

After checking them now in more detail and thinking about it I would say they do not help as much as I was hoping. They are sometimes more or less easy measurables and the other time they represent very hard to grasp complex things (e.g. disturbance regime).

cpfaff commented 8 years ago

By that I mean they do not solve the problem we have. They are rather of the same kind like what we already have.