(Note this article references my previous post entitled “Gustav and New Orleans: A Political Risk Assessment for the Republican Decision Maker” dated Saturday, 08-30-08, and “UPDATE: Gustav and New Orleans: A Political Risk Assessment for the Republican Decision Maker” dated Sunday, 08-31-08).

Events are rapidly unfolding.  According to the National Weather Service, National Hurricane Center Advisory Number 31:

• The center of Gustav is not poised to strike New Orleans when the storm makes landfall.

• New Orleans will experience at least tropical storm force winds, and perhaps hurricane force winds depending on whether the storm veers slightly east (note this information was derived from a procedure, whether visually or computationally, that the NHC labels as “experimental”).

• And, New Orleans, with a probability [90%, 100%] will experience a storm surge.

• Gustav, however will, with near certain likeliness, be a category 3 storm or less (>96% chances), with category 2 being the most likely storm intensity.  Since the storm will likely not hit New Orleans dead on and at landfall the storm will be weaker than Katrina (which was a category 3 storm when it hit New Orleans dead on), the effects of Gustav will most likely not be as bad Katrina.

Assessment Revisited

Because there is no question at this point that New Orleans is set to experience an appreciable storm surge due to the forces of Gustav, I revised the possibility tree to exclude scenarios E7 (i.e., Pr(A2) = 0; Pr(A1) = 1) and E1 through E3 (i.e., Pr(B1) = 0; Pr(B2) = 1) as follows:

All other probabilities remain unchanged (even if they are irrelevant given that their conditionals are no longer deemed plausible).  Thus, I revised my probability estimates for the remaining scenarios as follows:

Recall from my previous post that the political outcomes for scenarios E4 through E6 (which I leave unchanged) is as follows:

• E4: “-”
• E5: “-”
• E6: “++”

Now, with both consequence and likeliness in hand, the political risk profile charactering the uncertainty associated with near-term political outcomes can be illustrated as follows:

Summary of Results

Based on my revised risk profile derived from all assumptions implicit in the scope of this analysis, the most likely scenario is that the Republicans will benefit from Gustav, at least in the near term.  In fact, the analysis assigns odds of 19:1 in favor of the Republicans.  Perhaps this is the reason that Time Magazine is posing the question “McCain to Seize the Stormy Moment?“  But, the CCAN blog notes that Gustav is another example of extreme weather afflicting an nation in denial.  Perhaps the Democrats can spin this whole situation their way by emphasizing Republican denial of global warming, citing Gustav as “evidence” in support of the global warming is real hypothesis.  My particular analysis, however, does not look into this issue.

Now, given that my conclusions rely on fewer assumptions than with my previous analysis, I assign MODERATE CONFIDENCE to this analysis.  This is the highest level of confidence I can assign given the time I have available for analysis and my level of expertise in the subject.  I must remind readers here that the scope of this analysis is focused on the NOHPS, and does not consider the likeliness and political outcomes for some other non-New Orleans disaster (that could have been prevented by something designed and installed to prevent it).

I think the next post on this subject will come when Gustav is done.  Let’s see what happens…

Send article as PDF to

(Note this article references my previous post entitled “Gustav and New Orleans: A Political Risk Assessment for the Republican Decision Maker” dated Saturday, 08-30-08).

Revision of Likeliness Estimates

According to the National Weather Service, National Hurricane Center, Gustav is veering to the right (albeit slightly) of yesterday’s projected storm track centerline (see below).  Moreover, the National Weather Service has deemed it sufficiently likely that Gustav will significantly impact New Orleans that it has issued a Hurricane Warning to the city and points all along the Louisiana and Mississippi coastlines (and a little bit of Alabama too).  Accordingly, I revise my probability estimate for the event that Gustav will afflict New Orleans in an appreciable way (in terms of storm surge) as follows:

• Pr(A1) = 0.75 (increase from yesterday’s 0.50)
• Pr(A2) = 0.25 (corresponding decrease from yesterday)

Note that I did not use Bayes’ Theorem to update my probability estimates, but rather did a complete reassessment of the probability.  Also note I continue to entertain non-zero chances that the storm will not afflict New Orleans since the storm track envelope admits some degree of likeliness to the storm veering to the west.

Now let’s turn out attention to the time Gustav makes landfall.  According to the right-most figure above, the storm center will probably make landfall sometime between 7AM CDT and 7PM CDT on Monday, 08-31-08.  According to the maximum wind speed probability table (NWS advisory 28, dated 10AM CDT, 08-31-08), the chances that Gustav will be a category 3 storm or greater is [37%, 81%].  Since Katrina was a category 3 storm, lets assume that if Gustav is classified as a category 3 storm or greater at landfall AND the storm is one that is slated to yield an appreciable storm surge affecting the New Orleans Hurricane Protection System, then Gustav will be as bad or worse than Katrina.  Otherwise, it will not be.  With this information, I can revise my probability estimates as follows:

• Pr(B1|A1) = [37%, 81%] (probability that Gustav will be as bad or worse than Katrina)
• Pr(B2|A1) = [19%, 63%] (probability that Gustav will not be as bad or worse than Katrina)

Since the other events C1, C2, D1, and D2 have to do with the integrity of the NOHPS (which is assumed to have not changed in the last day) and the consequences to New Orleans in the event of NOHPS failure (also assumed to be unchanged), the revised probability for the scenarios E1 through E7 is as follows:

Revised Consequence Assessment

The press recently reported to Republican decisions that, though most likely pursued for unselfish and humanitarian reasons (as they should), can also be viewed as strategies to decrease the potential for unfavorable outcomes and increase the potential for favorable outcomes.  These are as follows:

• As stated by White House Press Secretary Dana Perino (cited by CNN and elsewhere), President Bush and Vice President Cheney will not attend the Republican National Convention in St. Paul Minnesota in person; instead, Bush may address the convention via satellite.
• As reported by CNN (link), RNC organizers are examining options to change the schedule of the convention, to include delaying opening ceremonies, turning the convention into a massive telethon, or postponing the convention altogether.

In the context of this study, these measures may very well decrease the severity of the resulting political fall-out in the event of a NOHPS failure, but perhaps not by much because after all, the NOHPS failed despite having three years to rebuild.  On the flip side, if Gustav does afflict New Orleans in an appreciable way (again with respect to storm surge along the NOHPS), and the NOHPS successfully protects the city, then the Republicans will surely emphasize the improved ability of the federal government (under the Bush administration) to prepare for and protect against natural perils.  In this case, more of the same might not be viewed as being all that bad with respect to domestic preparedness (and perhaps by extension, homeland security).  But now, in the event that Gustav does not afflict New Orleans, the political fallout is slightly in favor of the Republicans since they are taking preventive measures to protect the Gulf coast across the board, and you can’t deny that doing a good job at this by putting the people first (which I argue they will do) will garner some brownie points from the electorate.

Given the above, I revise my political outcome assessment as follows:

• E1: unchanged at  “–”
• E2: unchanged at “-”
• E3: unchanged at “++”
• E4: unchanged at “–”
• E5: unchanged at “-”
• E6: improved to “++”
• E7: improved to “+” (may change back to “N” if the Democrats can somehow display on a national stage that they are doing something too)

Given the revised likeliness and political outcome assessments above, the revised risk profile over the space of political outcomes is as follows:

Revised Results

Based on the revised risk profile above, we can draw the following revised conclusions:

• The most likely major scenario is that the Republicans will benefit from Gustav [86%, 93%]
• The must less likely scenario is that the Republican campaign will suffer from the effects of Gustave [7%, 14%]
• It seems at this point that Republicans will feel something given that the probability of a neutral or “no effect” outcome is near zero at this stage

From the Democrat perspective, one strategy to mitigate the Democrat risk is to do something that will gain them visibility on the national stage in relation to Gustav preparedness.  A visible substantial effort may serve to transfer some of the Republican benefits resulting from scenario E7 back to the neutral position (I doubt they can do much to pull any support in their favor for this scenario).  But even still, given the probability assessment being what it is, the ordering of the major scenarios will remain unchanged.

CAVEAT: Given the probabilistic nature of my assessment, the only verifiable portion of my assessment is whether or not the neutral outcome is realized.  But this requires me to stand by my assessment in light of new information (which I am sure to come across).  I suspect that I will find myself updating this analysis again sometime in the next day or so.

Note: as stated in my previous post, I assign LOW CONFIDENCE to this analysis given that I am not an expert in hydrodynamics, nor am I an expert in politics and political sentiment.

Send article as PDF to

At 5PM EST, Saturday, 08/30/08, the National Weather Service, National Hurricane Center released their estimate of probability of storm surge afflicting New Orleans sometime over the next 72 hours (see below):

Mayor Ray Nagin of New Orleans is labeling Gustav as the “mother of all storms.”  Apparently, New Orleans (the brownish colored area amidst the probability contour) has between a 40-60% probability of storm surge (it is hard to tell whether New Orleans is in the yellow, brown, or partial to both).  What was my estimate from the 2PM EST data?  My estimate was 50% that Gustav would result in an appreciable storm surge.  So, I suppose this is confirmation of my initial estimate (perhaps not, but it is nice to think so).  The next question is whether Gustav will result in storm surges as bad or worse than Katrina.  Based on what I am reading, I suspect so, but we shall see.

(note: this post references my previous post from just a few minutes ago)

Send article as PDF to