Rice: The Global Sliced Bread

To focus an entire project on rice may seem insane, and maybe it is. But rice is the new sliced bread, the famine food, the staple that feeds more than 50% of the entire world. As a crucial crop for low income populations, it is responsible for 20% of the daily calories for more than 3.5 billion people, “19% of global human per capita energy, and 13% of per capita protein in 2009” (“The global staple”). Not surprisingly, Asia is responsible for 90% of the world’s rice consumption, but the demand is increasing rapidly in other countries, becoming the “fastest growing food staple in Africa, and also one of the fastest in Latin America”(“The global staple”). Not only is rice an essential nutrient source to ensure food security in developing countries, it is also correlated with political stability. The shifts in rice paddy cultivation impact a much larger breadth of global issues than usually assumed.

One of the largest areas for rice production is the Rice-Jute-Tea Region in India. Included in this expansive region are the states of Arunachal Pradesh, Assam, and Meghalaya, along with many others(Mondal). While some states have seen a rise in rainfall in the past 30 years, Meghalaya and Arunachal Pradesh are among 7 Indian states that have shown “statistically significant decreasing trends in monsoon rainfall”(The Wire Staff). Global warming is most likely responsible for both of these trends. In this report I explore the fluctuations in precipitation in these three rice production regions to attempt to predict the annual rainfall in 30 years, in order to hypothesize how rice paddy production will be affected in these regions.

Looking at changes in precipitation is a useful way to understand how rice paddy agriculture will evolve because rice needs much more water than most other staple crops to be sustainable. The minimum average annual rainfall needed for paddy cultivation to be possible is 115 cm but the most desirable, and successful, conditions range between 175 to 300 cm per year (Chand). Currently the annual rainfall in the Rice Jute Tea Region is said to be around 180 to 250 cm (Mondal), mirroring the data I found stating that in 2015 average annual rainfall for the Arunachal Pradesh region was 276.8 cm, and was 247.1 cm for the Assam and Meghalaya regions.

I first used simple linear regression to plot the rainfall in mm. My model for the Arunachal Pradesh region predicted that the average rainfall would be 1370 mm or 137 cm for the year 2050, and in 2060 this number would drop to 113 cm, 2 cm below the minimum for paddy cultivation. For the Assam & Meghalaya region, the rainfall decline is much less rapid with 234 cm predicted for 2050, and 231 cm predicted for 2060. Both annual rainfalls for these cultivation sites are well within the desired range. To check the accuracy of the linear fit I calculated the RMSE of both plots. From below you can clearly see linear regression is not a good fit for either, especially to plot the Arunachal Pradesh linear regression as a negative r squared value indicates that a horizontal line would be a better fit than the model.

Then to try to improve the accuracy of my regression prediction line I used a linear fit of the log of the dataset. The fit was not much better than linear, producing similar r squared values of and rmse values. Finally, I fit the data to polynomial regression which although not an extraordinary model revealing week r squared or rmse scores, it was noticeably more accurate than the linear or linear log regression.

The predicted values from the polynomial regression plots revealed less of a decline in rainfall than shown in the linear fits, predicting 202.6 cm of annual rainfall in 2050 for the Arunachal Pradesh region, and 221.7 cm for the Assam and Meghalaya region in 2050. While these numbers are within the bounds to successfully cultivate rice paddy, the plot conveys a steep negative trend, indicating this soon will not be the case.

With a rapidly growing human population, an increased demand for rice production is inevitable. Some researchers even say that for approximately “every one billion people added to the world’s population, 100 million more tons of rice (paddy) need to be produced annually”(“The global staple”). The stark decline in precipitation in the crucial rice production regions Arunachal Pradesh, Assam, and Meghalaya , along with no promise of future rice paddy plots expanding into other areas, will likely result in a shortage of rice. If rice yields don’t keep up with global consumer demands, rice market prices can’t be stabilized, the affordability of this food staple will be threatened and millions of low income populations are at risk for malnourishment (“The global staple”). To truly understand the severity of this precipitation and paddy cultivation decline, I would need to delve further into other regression models to obtain more accurate prediction models. Are my rainfall predictions relatively close or will the world have more time before it has to worry about the future of rice?

Works Cited

Chand, S. (2014, February 08). Cultivation of Rice: Suitable Conditions Required for the Cultivation of Rice (6 Conditions). Retrieved December 11, 2020, from https://www.yourarticlelibrary.com/cultivation/cultivation-of-rice-suitable-conditions-required-for-the-cultivation-of-rice-6-conditions/25491

The global staple. (n.d.). Retrieved December 11, 2020, from http://ricepedia.org/rice-as-food/the-global-staple-rice-consumers

Mondal, P. (2014, February 22). 6 Types of Agricultural Regions of India. Retrieved December 11, 2020, from https://www.yourarticlelibrary.com/agriculture/6-types-of-agricultural-regions-of-india/20987

The Wire Staff. (2020, March 10). Rainfall Declining in Seven States, Analysis of 30 Years of Data Shows. Retrieved December 11, 2020, from https://science.thewire.in/environment/rainfall-declining-in-seven-states-analysis-of-30-years-of-data-shows/