Understanding mechanisms of delayed monsoon onset with warming
It is impossible to overstate the importance of monsoon systems for the Earth's climate both at the global and regional scale. These systems, which develop on all subtropical continents, including those of Southeast Asia, West Africa and North America in the northern hemisphere and Australia and South America in the southern hemisphere, are intrinsic elements of the tropical atmospheric overturning, namely the Hadley cell, and, as such, shape in fundamental ways the global cycles of moisture, energy and momentum. Their societal importance is even more dramatic: monsoonal rains sustain more than half of the human population in countries that have rapidly growing economies and that are critically dependent on the arrival of bountiful rainfall in the summer for agricultural planning, human consumption, energy generation and industrial development. The economies of these countries to these days strongly rely on agriculture. For instance, in India (affected by the South Asian monsoon, the largest scale monsoon in the Earth's atmosphere) agriculture accounts to 15% of the gross domestic product and employs two thirds of the Indian population. More than half of Indian agriculture remains non irrigated and agricultural practices have been developed to best exploit the remarkable regularity of the annual cycle, characterized by an alternating pattern of dry and cold winters and warm and wet summers (e.g., Kumar et al. 2004). In spite of, and in fact because of this remarkable regularity, agricultural practices are extremely susceptible to even small changes in the annual cycle of precipitation (e.g. Webster et al. 1998; Gadgil 2003; Turner and Annamalai 2012), including timing of monsoon onset and withdrawal, total season precipitation amounts and precipitation variability during the monsoon season.
With projected increases in population and pressure for food security, it is hence no surprise that understanding how monsoons will change with changing climate has been a major priority for the tropical atmosphere dynamics community (e.g., Bony et al. 2015). In fact, uncertainties in monsoon projections in the Coupled Model Intercomparison Project Phase 3 (CMIP3) and Phase 5 (CMIP5) general circulation models (GCMs) remain significant both globally and regionally (e.g., Turner and Annamalai 2012; Lee and Wang 2014). Efforts to better constrain and reduce these uncertainties include theoretically-based studies to advance fundamental understanding of monsoons (e.g., Biasutti et al. 2018), model intercomparison projects (e.g., Voigt et al. 2016; Zhou et al. 2016) and process-oriented studies (e.g., Pascale et al. 2017). Most of these efforts have however focused on projections of seasonal-mean precipitation intensity and extent, while less attention has been given to seasonality changes. And yet, seasonality changes have critical societal and economical impacts, as shifts in timing of monsoon onset/withdrawal and season length, for instance, affect the number of crop cycles and overall yield, even with no change in seasonal rainfall. Intriguingly, CMIP models do robustly project a delay in the timing of monsoon onset, with a redistribution of precipitation from early to late in the summer season (e.g., Sobel and Camargo 2011; Seth et al. 2011, 2013). A satisfactory explanation of this delay has yet to emerge, and it remains unclear to what extent in some monsoon regions this might be an artifact of biases in simulations of present-day seasonality (e.g., Pascale et al. 2017).
In the proposed work, we want to address this knowledge gap. Specific objectives include:
- Expose relevant mechanisms in a hierarchy of climate model simulations
- Relate these mechanisms to those mediating the observed year-to-year variability of monsoon onset in different monsoon regions. While achieving these goals will require more than the one year of requested funding, within this time frame we expect to be able to answer some critical societally relevant questions: With increased greenhouse gas concentrations, will monsoonal precipitation shift in time to a later start? Will the overall length of the monsoon season change to, for instance, a shorter season with increased mean precipitation? Are similar changes occurring in all monsoon regions?
The goals of this project are ideally aligned with the THOR center mission to reduce the risks and costs of natural hazards facing society. Billions of people in socially and economically vulnerable regions rely on monsoonal precipitation. It is imperative to better constrain and reduce the uncertainty associated with projections of monsoon changes with warming. Here, we specifically focus on one aspect of these changes: possible shifts and changes in precipitation seasonality. While having received relatively little attention in the literature, these changes might be of primary importance and might have critical impacts on management of water resources for agricultural planning and other economical activities in all monsoon regions.
Predicting riverbank erosion hazards in melting permafrost
(Lamb and Fischer)
Riverbank Erosion Hazards - Most arctic landscapes have permanently frozen subsoil, or permafrost, that keeps the land intact and habitable along coasts and riverbanks. However, the Arctic is warming causing widespread destabilization of banks and rapid rates of lateral river migration, which destroys infrastructure and threatens the livelihood of arctic communities (USACE, 2009; USGCRP, 2018). The problem is compounded because the Arctic has warmed twice as fast as the global average during the past half century, and warming is expected to accelerate in the future (Serreze and Barry, 2011). Bank erosion is so severe that it is the principle cause for displacement or relocation of entire communities (Bronen, 2013; Trainor et al., 2017). Public Radio International has reported on Alaskan villages as "America's first climate change refugees" (Wernick, 2015) due to bank erosion. Independent reports from the U.S. Army Corps of Engineers (USACE, 2009) and the US Government Accountability Office (GAO, 2003) identify more than 87% of the 213 Alaskan native villages are affected by accelerated erosion due to climate change. At immediate risk are homes, water lines, fuel tanks, schools, health clinics, sewage treatment areas, and air strips (USACE, 2009). Villages struggle to move their communities away from banks that can retreat up to tens of meters per year, as we witnessed firsthand during a Caltech expedition last August. Some villages (e.g., Minto, Eagle and Newtok) already have relocated entirely. In the recent Fourth National Climate Assessment Report (USGCRP, 2018, Ch. 24), an authoritative report delivered to the US President and Congress, bank erosion in Alaska is identified as one of the major and immediate hazards associated with a warming climate, and they highlight the disproportionate effects it has on indigenous communities as one of their twelve high-level summary findings. President Obama used erosion in Alaskan communities as his case-in-point on the impacts of climate change (CBC, 2015). The problem may be far worse than articulated in recent hazard and climate reports; accelerated bank erosion could be the tipping point to major morphologic changes of arctic rivers. Most large arctic rivers are meandering (e.g., Yukon, Mackenzie, Koyukuk) and permafrost is thought to be the primary agent for bank strength that allows meandering (Costard et al., 2003). The loss of permafrost could cause rivers to shift from a meandering to a braided state (Parker, 1976; Eaton et al., 2010). In this scenario, rivers are expected to widen by 100-fold or more, shallow substantially, and significantly reduce their ability to transport water and sediment. This would be an extraordinary change to the terrestrial landscape and a catastrophe for arctic communities, unlike anything we have documented historically.
We propose that existing models are missing a key component: sediment entrainment. In order for a riverbank to erode, the river flow needs to be sufficient to entrain the thawed bank sediment, in addition to melting the pore ice (Dupeyrat et al., 2011). Indeed, rivers outside of permafrost terrain can have relatively stable banks provided by the mechanical strength of grains (i.e., granular friction and cohesion) and vegetation (Parker, 1976; Braudrick et al., 2009; Eaton et al., 2010). However, we currently lack a model that couples sediment entrainment and
pore-ice melting to determine bank erosion rates. This is a major knowledge gap because river response to warming is likely to depend very strongly on whether the erosion rate is entrainmentlimited or ablation-limited.
The combined theory and experimental work will be the first of its kind to explore the transition from ablation-limited to entrainment-limited erosion in permafrost riverbanks. In doing so, this work will be the first major step in developing quantitative predictions of bank erosion in permafrost landscapes, and their sensitivity to warming. This is a pilot project that presents an opportunity to make a fundamental contribution while addressing a major hazard of immediate concern and of global interest. This proposal will fund the first experiments in the new permafrost flume in the Caltech Earth Surface Dynamics Laboratory. We have already purchased and installed the chiller, but experiments have yet to be conducted, and the work is presently unfunded. Ultimately this project fits within a larger vision of a new major research direction for Lamb in permafrost landscapes, including the role of bank erosion in liberating frozen plant material that contributes to greenhouse gases with Fischer. We believe the preliminary data from this pilot study will show proof of concept and demonstrated success that is necessary to establish ongoing funding in acrtic programs in NSF and DOE that are new to us.