Climate change is a major cause of concern in the scientific community all over the world. There is much empirical evidence that the global temperature is rising along with various other adverse effects, and human activities like industrialization is to blame. To curb these, it is necessary to adopt mitigation strategies immediately. However, to convince policymakers and other stakeholders to do so, it is necessary to prove that climate change is indeed happening, by attributing specific weather phenomena to it. Not many studies of this kind have been done in the context of India, which is trying to industrialize rapidly while there are millions vulnerable to climate change.
A research project in Indian Institute of Science, Bangalore has recently tried to bridge this gap, by studying the variations of maximum and minimum temperatures, annual as well as seasonal. They have focussed on the all-India average temperature, as well as those over certain homogeneous regions of India. The main conclusion drawn from the study is: while it is difficult to attribute the temperature variations to any specific cause, the temperature fluctuations are definitely not ‘normal’.
The study has focussed on the monthly temperature data for the period 1950-2005. It started by noting the maximum and minimum temperatures for every year, and also for every season within each year. These temperatures were averaged over the entire country, as well as over special “homogeneous regions” of the country across which temperature variations are relatively low- such as the East Coast, West Coast, Western Himalaya etc. The various time-series of maximum and minimum temperatures, are processed and trends are extracted. These trends are then compared to corresponding trends obtained from simulations. Such simulations are done using various climate models that have been developed over the years by various research groups, and have been compiled together by an international project called Coupled Model Intercomparison Project-5 (CMIP-5). These models can simulate climate variables such as temperature over a period and region under normal conditions, as well as under different kinds of “forcings”- i.e. changes in the Earth’s energy balance due to natural or human economical activities.
For detection and attribution of climate change, it is necessary to quantify climate responses to forcings. “Fingerprint” is a compact mathematical representation of climate responses, which can be computed from simulations of past climate. This study first compares each of the different time series measured over the various regions and time-periods under consideration, against the simulation outputs of different models under perfectly natural settings (i.e. pre-industrial control scenario) using the fingerprint technique. For both maximum and minimum temperature, observations are compared with the climate model's simulation of pre-industrial control. Some statistical tests are carried out, and those cases where the observations vary widely from the model simulations are identified. In these cases, the observed time-series are further compared against model simulation outputs that consider various kinds of natural and human-induced forcings.
It turns out that the minimum temperature observations differ quite significantly from the model simulations, in natural conditions. According to Sonali Pattnaik, one of the researchers, “It is noticed that the emergence of observed trends is more pronounced in case of minimum temperature compared to maximum temperature”. This indicates that the observed variations of minimum annual and seasonal temperature are beyond the limits of natural variability in the climate. Under various conditions of natural and human-induced forcing, the observations generally agree with the model simulations, but some disagreements are also there. The researchers judge that confident attribution of the variations to any particular cause is not possible, but point out that they are not natural either.
This observation may necessitate various policy changes. Sonali Pattnaik points out, “Decisions related to hydraulic structure design, water supply planning and management based on stationary climate are no more valid”.
About the paper: “Detection and attribution of seasonal temperature changes in India with climate models in the CMIP5 archive” has been recently accepted for publication in the Journal of Water and Climate Change.
About the authors: The study group consisted of P. Sonali and Prof. D. Nagesh Kumar from Department of Civil Engineering, IISc.