In case you missed the 2018 fall meeting AGU session on compound events (NH23A) or just want a short recap, here we summarized the oral presentations.
The research field of compound events looks at many different perils. This session covered flooding, hurricanes, earthquakes, wildfires, wind but also how such perils may concurrently or consecutively combine. In order to genuinely capture the impact of such phenomenon, researchers presenting in this session focused on novel multivariate methods and/or extensive modelling to analyse these complex systems, either at local or global scales. Compared to the same session at EGU earlier in 2018, there was more focus on the impacts rather than the drivers of compound events.
This session covered flooding, hurricanes, earthquakes, wildfires, wind but also how such perils may concurrently or consecutively combine.
Amin presented an interesting case study of the Houston Ship Channel during hurricanes Harvey and Ike. Through hydrodynamic model simulations he showed that local runoff from precipitation added to the duration and depth of flooding during these hurricanes.
Dirk presented an impact based assessment of compound discharge and surge flooding at the global scale. He developed an indicator that assesses the change in water levels during peak events at the river mouth due to surge – discharge interactions. Using this indicator he found clear spatial patterns of compound flooding in both the magnitude of change as well as the timing of the largest changes yearly. The full presentation can be found here.
Kathleen (Kate) White – invited speaker
As a water resource management agency, the United States Army Corps of Engineers (USACE) has to address various impacts related to the water cycle and its change. Kathleen provided a broad overview of the current USACE activities related to compound events and cascading impacts. Recent analyses have highlighted the occurrence and potential threat from these events on water systems. Opportunities for improvement for such agencies when dealing with compound events were identified. The full presentation can be found here.
Amir Aghakouchak – invited speaker
Dr. AghaKouchak was an invited speaker in this session. He discussed how many studies have shown a substantial increase in ground-based observations of extreme rainfall events, hot spells and heatwaves. Nonetheless, only few studies address the changes in compound and consecutive events. Dr. AghaKouchak showed the importance of addressing those disasters where a single event may not present an extreme, but a combination of events occurring simultaneously or shorty after each other can become an extreme event, as is for example the case with drought-heatwaves.
Marleen de Ruiter
Current state-of-the art regional and global models and their outputs do not allow for a thorough representation and analysis of consecutive disasters and their impacts. Marleen underscored the relevance of improving our understanding of consecutive disasters by showing a global analysis of consecutive disasters consisting of historic earthquakes and typhoon events, and a local analysis of historic earthquakes, typhoons, volcanic eruptions and floods in the Philippines.
By analysing auto-correlations at different time-scales, Sonu showed that memory in meteorological and hydrological processes impacts flood waves. Peak discharges at the Rhine River decrease significantly by removing autocorrelation from meteorological forcing data up to time scales of 30 days. It is therefore important to account for the memory of soil-moisture, snow accumulation and antecedent hydro-climatic conditions.
By analysing disaster sequences in great detail, Robert disentangled the systemic correlation of fires following a disaster. While fires can be expected under certain circumstances such as strong earthquake shaking in wooden building areas, it can also be the result of long chain of completely independent events. This highlight the complexity of integrating such hazard in risk modeling.
Katherine (Katy) Serafin
Return water level design events used to derive flood adaptation strategies are often derived from individual rather than multivariate processes.
Katy showed how such approach may become limited in coastal urban areas where compound forcings are present. Her analysis of the water return levels within a selected river reach in the San Francisco Bay presented the relative importance of each driver to the design water level. This suggests that interventions purely based on univariate forcings may be counterproductive for protection from multivariate processes.
This summary blog was written by Marleen de Ruiter, Anaïs Couasnon and Dirk Eilander.