By: Kieran Hunt & Arathy Menon
These days, when a weather-related catastrophe occurs, one of the first questions raised in the aftermath is “did this happen because of climate change?”. Because of the stochastic and chaotic nature of weather, it is all but impossible to determine whether a single event was caused by climate change. There are, however, experiments that we can do to figure out whether climate change makes a certain type of event more likely, or for a given case, to what extent it has modified the impacts.
Our study explores the second of these options in the context of the devastating Kerala floods of 2018. During mid-August of that year, a monsoon depression passed unusually far south over the Indian subcontinent. This, in turn, excited the moist monsoonal westerlies, causing very heavy rainfall when they struck the mountain range that runs along the southwest Indian coast – the Western Ghats. The deluge fell mostly over Kerala, which had been saturated just several weeks earlier from rains associated with another low-pressure system. The reservoirs rapidly hit capacity, dams were opened state-wide, and the resulting flooding killed 483 people and displaced over a million more.
Figure 1: Average rainfall over 15-17 August 2018 (computed using data from NCMRWF). Also shown are the tracks of the precursor low-pressure system (6-9 August) and monsoon depression (13-17 August). The border of Kerala is shown in thick black.
Kerala lies mostly over the ecologically fragile Western Ghats and has a complex topography with the Arabian Sea to the west and mountains to the east. It also receives a large amount of rain with an average of about 300 cm during the monsoon season. About 50 major dams in Kerala provide water for agriculture and hydro-electric power generation. As a result of the torrential rains in August 2018, the authorities had to open the sluices of 35 of these major dams as they reached maximum capacity.
Figure 2: A photo showing the flooded Periyar river, submerging the surrounding areas during the August 2018 flood (Source: The Hindu).
So, how do we probe the role of climate change in all this? We set up three experiments, using a technique called “pseudo global warming”.The first, a control, is a simulation of the 2018 Kerala floods as they happened using a regional weather model (WRF, with coupled hydrology to allow river simulation) forced at the boundaries with ERA-Interim reanalysis data. We use the control experiment to verify the model is working correctly (for example, by checking the simulated rainfall looks close to observations) and as a benchmark against which we can judge our other two experiments. For the first of our two perturbation experiments, we “subtract” the effect of observed global warming by using output from the CMIP5 pre-industrial experiments to adjust our boundary conditions – in essence keeping the high-frequency information responsible for the floods and modulating the low-frequency information that describes the background climate (e.g. large-scale changes in temperature and humidity). The second perturbation experiment uses the same method to “add” projected global warming in 2100 from the CMIP5 RCP8.5 experiments. Thus, our three experiments describe how the floods would look like in the current climate (control), a climate where no human-induced global warming takes place (pre-industrial), and a climate where much more global warming takes place (RCP8.5).
Results show that the rainfall affecting Kerala during August 2018 would have been 18% greater had human-induced climate change never occurred; in contrast, it would be 36% higher in the 2100 future climate. The first result seems counterintuitive at first glance: the world has warmed considerably since the pre-industrial era, and that warming brings with it a lot of additional moisture, so we would naively expect more rainfall, not less. What’s going on? Well, another result of climate change (both observed and projected) is a weakening of monsoon depressions – and in this case, that weakening has a stronger effect on the rainfall than the increase in humidity. This tug-of-war changes hands, dramatically, in the future climate experiment as the moisture increase easily overwhelms the weakened dynamics, which you can see in Figure 3.
Figure 3. Relative contributions to changes in moisture flux from changes in moisture (left column) and winds (right column).
How would this change in rainfall have affected the reservoirs and rivers? To answer this, we need to use a hydrological model that takes information from our weather model (e.g. rainfall, winds, temperature) and computes the response of local rivers and groundwater. Perhaps unsurprisingly, changes in the average river discharge over Kerala are almost identical to the changes in precipitation. However, given the highly variable Keralan topography, local responses to the climate perturbations can vary significantly. It’s beyond the scope of a blog post to go through each reservoir individually, so let’s focus on the largest one: Idukki. Built from nearly 500,000 cubic metres of concrete, the Idukki reservoir is responsible for over a quarter of Kerala’s total freshwater capacity. Figure 4 shows the modelled inflow rate and storage for the three experiments, with observational data for comparison. The model performs well, with simulated storage closely matching observations (phew!), at least until authorities opened the floodgates in mid-August. The most interesting take-away, however, is the gap between the respective orange lines and the dashed grey line – this represents the additional capacity that the reservoir would’ve needed to prevent flooding, and the minimum amount of water that would end up inundating downstream parts of Kerala. In the control experiment, this excess amounts to 589 million cubic metres of water in the control experiment, but 852 million in the future climate experiment, an increase of 45%. Other major dams show broadly similar patterns, although the effect of the future climate worsens significantly towards the south of the state.
Figure 4: Modelled inflow (blue:control; grey:pre-industrial; red:future) and storage (orange solid:control; dashed:pre-industrial; dotted:future) for the Idukki reservoir system. For comparison, black crosses show the daily observations of storage, and the grey dashed line shows the stated maximum capacity of the reservoir.
Summarising, the 2018 Kerala floods were likely made less damaging by climate change, as global warming has weakened monsoon depressions. However, if they were to happen again in a future climate (RCP8.5) scenario at the end of this century, the effect of increased tropical humidity would far outweigh the weakened depressions, likely resulting in a significantly more catastrophic scenario.
References:
Dash, S. K., J. R. Kumar, & M. S. Shekhar, (2004). On the decreasing frequency of monsoon depressions over the Indian region. Curr. Sci., 86, 1404-1411, https://www.jstor.org/stable/24109213?seq=1
Dee, D. P., and Coauthors, (2011). The ERA‐Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137(656), 553-597, https://rmets.onlinelibrary.wiley.com/doi/full/10.1002/qj.828
Gochis D. J., Yu W. & Yates D. N. (2014) The WRF-Hydro model technical description and user’s guide, version 2.0. Tech. rep., NCAR
Hunt, K. M. R., & A. Menon, A. (2020). The 2018 Kerala floods: a climate change perspective. Climate Dynamics, 54, 2433-2446, https://doi.org/10.1007/s00382-020-05123-7
Mitra, A. K., I. M. Momin, E. N. Rajagopal, S. Basu, M. N. Rajeevan, & T.N. Krishnamurti, (2013) Gridded daily Indian monsoon rainfall for 14 seasons: Merged TRMM and IMD gauge analyzed values. J. Earth Syst. Sci., 122(5), 1173-1182, https://www.ias.ac.in/describe/article/jess/122/05/1173-1182
Prein, A., R. M. Rasmussen, K. Ikeda, C. Liu, M. P. Clark & G. J. Holland, (2017) The future intensification of hourly precipitation extremes. Nat. Climate Change, 7, 48–52, https://doi.org/10.1038/nclimate3168
Sandeep, S., R. S. Ajayamohan, W. R. Boos, T. P. Sabin, & V. Praveen, (2018). Decline and poleward shift in Indian summer monsoon synoptic activity in a warming climate., Proc. Natl. Acad. Sci. U.S.A., 115(11), 2681-2686, https://doi.org/10.1073/pnas.1709031115