Dear Anissa Frini, Sarah Benamor and Bruno Urli

I have read your paper

Temporal MCDA Methods for Decision-Making in Sustainable Development Context

My comments are:

1- In page 3 you say “These methods enable the simultaneous consideration of conflicting criteria as it occurs in a real-world problem under sustainability imperatives”

I am afraid I disagree because most MCDM methods consider criteria individually; there is not a simultaneous consideration of them

“most articles surveyed in [2] did not investigate the long-term perspective related to sustainable development”

Very true

“The impacts of decisions are uncertain”

In sustainability, at least, I don’t think that the impact of decisions is uncertain, if potential consequences are analyzed, for example, in eliminating certain aspect to improve another.

Have a look at this article published by a Jakov W, Frank published in ‘Defenders of wildlife’

“In a little over a decade – from 1914 to 1926 – we eliminated 136 wolves that lived within the park. This was before we understood the connectedness of wildlife – before we knew the value of intact ecosystems. Before we knew about climate change and biodiversity and loss. Wiping out an apex predator in the park turned out to be a major mistake”

“Seventy years without wolves changed Yellowstone – songbirds left, elk and coyotes became overpopulated and beavers disappeared. Elks overgrazed the land and trees, such as willow and aspen. Without those trees, songbirds began to decline. Beavers could no longer build their dams, and streams began to erode and degrade the conditions willow trees needed to grow. Without beaver’s dams and the shade from trees and plants, water temperatures were too high for cold water fish.”

Remediation: Wolves were reinserted in Yellowstone Park, and the original decision reverted.

As another case, it was NOT UNCERTAIN the impactregarding stopping the daily dumping of New York waste into de ocean.

In my opinion, there are two different problems that must be addressed:

1- To decide what would be the best technique to improve environment, something that can be approached by MCDM, and

2- Which could be the consequences and impact of such decision

I believe that the most outstanding example of the divorce between options to ameliorate a problem, and the consequences of each one, is the generation of electricity using nuclear fission reactors

The option:

1- Generate electric energy using nuclear fission plants, as was decided in early 50s to have an abundant energy.

Did anybody thought what to do with the spent and highly radioactive uranium rods?

Yes, bury them in deep salt caverns or keep them for centuries in water pools until they degrade.

Which was worse, the cure or the disease? And it can’t be reversed.

Did anybody meditate about the consequences of storing spent rods?

Apparently, neither thought about the risks involved, nor about the consequences

And the cost, the risk of transportation?

We are leaving to future generations to solve the problem (putting the dirt under the carpet…).

And all of this was known in the 50s. Therefore, after than more 70 years we don’t still know what to do!

2-” These methods enable the simultaneous consideration of conflicting criteria as it occurs in a real-world problem under sustainability imperatives”

Sorry, as far as I know there is only one method that can do that: Linear Programming

I would say that 99% of MCDM methods, analyze alternatives by one criterion at the time, and thus, assuming that other criteria do not have any influence in that analysis.

3- Page 4, “Criteria C1,C2 … ,CM are assumed to be independent”

If they are independent, what you are doing is a simple addition, not getting the result due to interactions

4. I agree with Figure 1, because in my opinion this is the way to consider temporality, and as you say before, it allows for working with different criteria and values in each period. Also, consider that alternatives can also change in number and in type, something that this system allows.

5- What is what you call ‘preference relations? Are they the results of pair-wise comparisons?

If they are, what are the based on? Personal appreciations? Not very scientific indeed!

Why are you comparing actions? As I understand it, in this problem you do not aim at determining importance. In addition, measures that you take in period 1, say reforestation, MUST continue in the next periods, if you are aiming at sustainability, even if you are logging. That is, there is a precedence here from period ‘n’ to period ‘n+1’. From this, to period (n+2), and so on.

The MCDM method must be instructed to consider it, if not, the software may delete in period (n+1), what was chosen in period (n)

6- In page 6 “Step 1.1: For each period t and for each pair of actions (ai, ak) , compute the concordance index C …’

What for? Maybe I am mistaken or don’t understand your problem, but I believe that you are looking for the best set of alternatives that maximize sustainability, that is, in years to come, and using resources also for our present needs. But, in my opinion, and since this problem is dynamic, we can’t apply the same rules and procedure used for choosing for instance a best location. We need to develop a set of alternatives that remain constant along time, and subject to changing criteria and perhaps a new alternative.

7- Page 7 “Actually, the selection of sustainable forest management options should arrive at a balance between biodiversity, soil and 7 Temporal MCDA Methods for Decision-Making in Sustainable Development water conservation, forest productivity, socioeconomic benefits, and the population’s values and needs”

I agree in a 100%, however it appears that the paper forgets that forest productivity, that is logging, produces a highly undesirable effect, erosion, that is, an externality. Where is it considered here?

How do you reach a balance if you ignore it? You can see it in British Columbia where the beautiful mountains appears with ugly patches of sterile land, courtesy of logging.

8- page 9 “The AHP method was used to model the preferences in terms of criteria weights”

Therefore, you are using the intuition of a DM or even of a group, to compromise 150 years?

“Relative importance of periods?”

I believe that all periods have the same importance

9- I think that the issue of temporality as you say, has not be properly addressed. It is uncertain, of course, but we can fix a target, albeit we don’t know if it can be accomplished or will be changed in the future. It appears that we need to aim at an indefinite future, but with a clear target, like to breath clean and pure air, abundantly fresh water, no waste, not noise, abundant wild life, plenty of forests, etc.,

In my humble opinion, when we address subjects that are well in the future, say for 2050, 2075 and 2100, we can’t use simulations or guessing, mainly because we don’t know what new tools for MCDM there will be in the future, therefore, we are constrained to use the tools we have now.

Just try to imagine how a scientist in 1890 could plan for the future, when practically everything that was top of the line in science at that time (electric light, railways, telephone, etc.), are completely different, in every field.

For this reason, to plan and pave the way for the future we only have a tool: Just analyze the trends, and mainly, establish goals, objectives or targets to achieve. In sustainability, we can establish objectives like: Any type of pollution reduced to near zero, generate electric energy using only renewable sources, increase reforestation, reduce externalities, reduce waste, etc., and we can put a value on those goals, not by guessing but supported in historic information and trends.

We have then an advantage by knowing the objectives or targets, and because we are looking for the best. Perhaps in 2035 a researcher finds that the goal was modest or very ambitious, and then, he can correct it, but the purpose of the target does not change, because our survival depends on it; circular economy is an example, forestation is another, as well as water, etc.

‘Most articles surveyed in [2] did not investigate the long-term perspective related to sustainable development in the year 2000

It is understandable; assume a researcher in transportation in 1950 trying to establish a goal for 2000. Probably he would envision higher speeds, large cars, more comfort, more powerful engines, better suspension, etc., lower fuel consumption was not even considered. Sustainability? It was an unknown concept. Electric car? Yes, as were the first cars, but in 1950 there was not lithium batteries our fuel cells

Of course, there is no guarantee that in 2050 we will reach the objectives for that year, but most probably we could be nearby, because we are more conscious and have tools than in 1950 were not even dreamed (computers, new sources of energy, better transportation, etc.). What may change are the ways to get it, for instance building nuclear fusion stations and closing nuclear fission utilities, decommissioning fuel-fired power plants, and developing sources such hydrogen cells and solar and wind sources of energy.

This is one of the most interesting articles related to MCDM I have read, and most important, with far reaching consequences. Really, the authors should be congratulated for bringing so important a subject, albeit, in my opinion, their article is too basic and incomplete and using a complex and arguably wrong method, based on personal opinions and not on research and investigation.

I hope my comments will help

Nolberto Munier