October 15, 2020: IRES Student Seminar with Balsher Sidhu and Luis Felipe Melgarejo Perez

IRES Seminar Series

Time: 12:30pm to 1:30pm (every Thursday)

Via Zoom

View video.


Machine learning methods for examining the impact of climate variability on crop yields in India

Due to the strong dependence of agricultural productivity on climate, predicting crop yields as a function of climate variability has been a topic of extensive research over the past many decades. However, most previous studies have utilized seasonal average climate values in their statistical models, largely ignoring intra-seasonal climate variability. Even though the latter has been demonstrated to have disproportionately strong impacts on crop yields, it has largely remained unaccounted for in statistical crop models and global yield estimates. To address this research gap, Balsher is developing improved crop yield statistical models with dedicated variables for intra-seasonal climate variability, using India as a case study. In this talk, Balsher will discuss a part of his research showing the advantages of machine learning methods over more traditional regression techniques for predicting crop yields as a function of climate variability.

Balsher Sidhu

IRES PhD Program


Balsher grew up in Punjab, India, a state often called the bread basket of the country. His formative years in the midst of intensive agricultural activity have played a prominent role in determining the topic of his doctoral research, for which he is analyzing the relationship between climate and agriculture in India. More specifically, he is building statistical models to quantify the impact climate variability (both intra-seasonal, and long-term climate change) has on crop yields across the country. Balsher is co-supervised by professors Milind Kandlikar and Navin Ramankutty, and is funded by UBC’s Four-Year Fellowship and the NSERC Vanier Canada Graduate Scholarship.