2 November 2022

The role of atmospheric pressure in landslide triggering

Posted by Dave Petley

The role of atmospheric pressure in landslide triggering

A few years ago there was considerable surprise when high resolution monitoring of the Slumgullion landslide in the United States found that atmospheric pressure played a role in controlling movement.  There has been some interest in this behaviour since – for example, it has not been clear as to whether Slumgullion is an exception. Is the stability state so marginal in this case that tiny fluctuations in inter-grain pressure affect stability – or has this type of response been missed because of the low resolution of most monitoring?

A new open access paper in the journal Natural Hazards and Earth System Science (Pelascini et al. 2022) has investigated this phenomenon in more detail.  The test site was in Taiwan, which I have always considered to be the world’s greatest landslide laboratory (as well as an amazing place to visit).  The weather conditions that drive the investigation were typhoons, which are really interesting because they bring both extreme rainfall and comparatively large fluctuations in atmospheric pressure.  The researchers have used a model to explore the slope response to changes in pressure, driving by measured changes in pressure from the typhoon events.

The diagram below, from Pelascini et al. (2022), illustrates the potential drivers of the landslides.  The graph on the left shows peak rainfall intensity for a range of typhoons in the study area plotted against the atmospheric pressure drop, with three key typhoons highlighted.  On the right, the time series for the change in atmospheric pressure and the rainfall is shown for the three typhoons and a synthetic event.

(a) Typhoons over Taiwan sorted as a function of their maximum rainfall intensity and atmospheric pressure drop. The hydrological context is represented by the colour scale, showing the cumulated rainfall over the 6 months before the event. (b) Time series of rainfall and atmospheric pressure changes of the three typhoons and the synthetic event created from the average of every typhoon.

(From the paper): (a) Typhoons over Taiwan sorted as a function of their maximum rainfall intensity and atmospheric pressure drop. The hydrological context is represented by the colour scale, showing the cumulated rainfall over the 6 months before the event. (b) Time series of rainfall and atmospheric pressure changes of the three typhoons and the synthetic event created from the average of every typhoon. Figure from Pelascini et al. (2022).

.

When this data was fed into the landslide model, interesting results merged.  Not surprisingly, the rainfall infiltration was able to induce large changes in pore water pressure, which are of course important for stability.  However, the effects were delayed by hours or even days due to the time required for pressure change to diffuse through the soil.  On the other hand, the atmospheric pressure changes were much smaller, but the impacts on the soil were instantaneous.

Pelascini et al. (2022) have made an interesting observation in that typhoons occur in or after the wet season in Taiwan, when slopes are already saturated.  I would be interested in understanding the data that supports this key assertion.  But, if so, then something else must finally trigger the landslides in the typhoon event.  Their modelling suggests that this might be the atmospheric pressure drop.

This is a fascinating result.  Of course, in the real world, this is a complex phenomenon, and the authors are careful to note the need for more research.  In a typhoon, triggering will be controlled at the first level by rainfall – the slope will not fail until the pore water pressures are high – but Pelascini et al. (2022) have highlighted that the factor that might finally induce failure is the reduction in atmospheric pressure as the eye of the storm passes over or close to the slope.  In a sense, the atmospheric pressure is the straw that breaks the camel’s back.

There is one other observation that is interesting in the paper.  The models suggest that storm triggered landslides tend to initiate at the toe of the slope.  This has been postulated previously, but it is good to see further evidence.

Overall, I think that this study is a really fascinating contribution, providing further evidence of the complexity of landslide initiation.  It should be the trigger (if you’ll excuse the pun) for a range of studies involving slope scale monitoring and modelling, as well as landscape-scale investigations. I shall look forward to seeing further work in this area in the coming years.

.

Reference

Pelascini, L., Steer, P., Mouyen, M., and Longuevergne, L. 2022. Finite-hillslope analysis of landslides triggered by excess pore water pressure: the roles of atmospheric pressure and rainfall infiltration during typhoons. Natural Hazards and Earth System Sciences, 22, 3125–3141, https://doi.org/10.5194/nhess-22-3125-2022.