BP has already blogged on the statement by Dr. Smith that the Irrigation Department should have released water earlier in the rainy season – see here and here – and then the level of water in the dams, water flowing into the dams, and discharge of water from the Sirkit and Bhumipol Dams. However, what about an alternative view?

Richard Meehan is an Adjunct Professor of Civil Engineering at Stanford University. He has sent BP some notes on the flooding. To save everybody from googling his background. Here is his bio from his website:

Richard Meehan is an engineer with degrees from MIT and Imperial College, University of London, and is an adjunct professor at Stanford University. He specializes in environmental and engineering problems of river valleys, including flooding, subsidence, and other hazards associated with urbanization.

BP: He has been quoted by AP and San Francisco Chronicle called him a “flood-control expert” and has also acted as an expert witness in court in the US.

Now, if you are thinking to yourself, okay, why should we listen to some westerner who knows nothing about the Thai dams. From his website:

Following service with the U.S. Army Corps of Engineers he began designing dams and levees in Southeast Asia in the early 1960s,

Construction Engineer, Royal Irrigation Department of Thailand, 1963-1965;

BP has made some slight changes for grammar and style and added emphasis (some images also not included due to image quality when trying to transfer):

Problem statement
The Bangkok flooding of 2011 occurred following a series of heavy summer and fall monsoon rains in Northern Thailand. This resulted in the usual late-fall flooding of poorly drained lands on the lower Chao Phraya floodplain, including newly developed industrial estates. Large damages distributed over many social and economic sectors resulted including extensive residential areas, both rich and poor, and industrial facilities supported by foreign investment and managed by modern techniques of risk management heavily dependent on insurance.

Though the damages were great, the flood itself may not have been an exceptional natural event, and it’s impact was certainly attenuated if not eliminated by the presence of dams in the upstream watershed. Last-minute efforts to spare the CBD and tourist areas of Bangkok succeeded, but some say at the expense of seriously aggravating the flooding in residential and industrial areas.

The future impact of flooding, almost certainly of volumes more threatening than the 2011 flood, will depend on future risk allocation policies by the government. Development of mitigating floodway and other floodplain mitigation programs will be costly. Future allocation of dam storage would increase the risk of water shortages. On the other hand failure to mitigate the hazard will not only bring suffering to the people and economy of the tradi- tional country but could lead to loss of insurance availability and flight of foreign investment which has created a modern Thailand.

Severity of the 2011 flood as a natural event

Based on a number of factors, including 2011 rainfall, July to December, Chao Phraya discharge at Nakhorn Sawan, and history of past floods, the 2011 flood (red square) appears to be an event with about 30 year return period.

Estimating the magnitude of a 100 year event, which seems more realistic “design flood” for planning purposes depends on whether traditional Gumbel-type techniques or a newer pro- posed “Power Law” technique is used. The latter method has been applied to the Mae Chaem River (part of the Ping watershed). It suggests a possibility of an enormous Chao Phraya design flood of about 75 billion cubic meters (versus about 35 billion cubic meters for the 2011 flood). On the other hand he traditional Gumbel approach suggests that the 100 year event is only modestly larger than the 2011 flood, perhaps on the order of 40 billion cubic meters. Depending on which philosophy is accepted, the difference in design of remedial measures to protect against future flooding is tremendous. And yet important decisions must be based on such highly uncertain estimates.

CAPTION: Different prediction philosophies lead to different predictions for the 100- year flood. Vertical is flood size (July- Dec) and horizontal axis Return Period. Black point is forecast by “Power Law”; red point is 2011 flood; blue is data from 1956-2001.

Although it is common to speak of the threat of an additive threat due to global warming, its additive effect is reportedly only about 3 percent. Even that is uncertain, there are data suggesting that increased rainfall and flooding are not a likely result of global warming in Southeast Asia. Uncertainty of predicting events such as the 100 year flood are more important.

The role of upstream dams
The Bhumipol and Sirikit dams were originally built mainly for retention of wet season water for irrigation and other uses in the dry season, and have consistently been described as water conservation projects, without reference to a flood control function, in many studies including those by international agencies and researchers over the years. This view of the proper function of the dams was reinforced by an ominous
trend in the early 1990s pointing to a drying up of the Chao Phraya River, Thailand’s “life- line”.  More recent evidence continues to show that global warming is reducing river flows in Southeast Asia.

Source: Bangkok Post

[BP: Although, as correctly noted in PostBag,

This terminology is incorrect. The correct names of the curves are ''Upper Rule Curve'' and ''Lower Rule Curve''. The word ''rule'' in this context means ''control'' or ''governing''. That's why the levels vary with seasons of the year. If the curves represented the maximum and the minimum storage capacities, they would simply be horizontally straight, as ''capacity'' is constant and does not vary with the seasons.]

Allocating part or all of the storage capacity of these dams to flood control would reduce or eliminate the primary function of retention for dry season use of increasingly scarce water. The “Rule Curves” used by RID to make decisions about reservoir discharges do not call for significant use for flood control. Discharging water in the reservoirs early in the season say July or August increases the risk that in years where the rains fall off late in the season the reservoirs will not fill and fail to meet the demands, mainly agricultural, below. These demands chronically exceed the capacity of the reservoirs. During the 2011 flood there was hindsight criticism of RID for not discharging water earlier, say in July, when it was “obvious” that the rains were continuing. But the operation of the Bhumipol dam as shown in the figure was consistent with the primacy of water conservation over flood control. Dropping the water in July or August to levels at or below the indicated minimum could have ended the year with a shortage as in 2009 and 2010. Only in retrospect does it appear that “they should have known the reservoir was too full” as some experts, including water engineers, have claimed.

My argument here, which amounts to a defense of RID operations, is open to various objections. The major shift in Thailand’s economy from growing rice to manufacturing disk drives and the accompanying population shift from village to suburbs (in low-lying areas) change the
balance of risks in a way that favors more flood protection. A half century of hydrological data, new theories on possible extreme events, better weather forecasting in the 30-60 day range, and better analytical tools allow for some fine tuning, optimization, and democratiza- tion of water management. Emergent notions of “Black Swan” events such as dam failures or flood-assisted flu epidemics raise new alarms. The Discovery Channel turns its prognostications from the desertification of Thailand to drowning of Atlantis. New policies will emerge, the Rule Curves may and perhaps should change. Resulting new policies may prove to be very successful, as was the case several years ago when Thailand faced the flood-related problem of ground subsidence (more on that later).

Studies in progress

Alternative dam management.
What difference in the impact of 2011 floods if Bhumipol and Sirikit dams had been been operated differently? Model study.

California’s floods: relevant examples.
Compare California’s central valley (eg Sacramento). Floodways, levees, changing California laws on compensation for flood loss.

Bangkok’s CBD: Why no flood?
Look at topographic history of Bangkok with the idea that there may be natural protection in the CBD that favors that area over the suburbs (contrast with the idea that protection was at the expense of suburbs)

BP: If BP understands the argument correctly, RID were simply operating within the rules they had for many years and the the extraordinary above normal rainfall this year caused the problem. Hence, it is more a system problem with the rule curves than RID making a mistake/losing their nerve.