An assistant professor in the School for Environment and Sustainability at the University of Michigan Meha Jain presented her research on ways to sustainably increase agricultural production, specifically in India  on Jan. 23 at Bren Hall. 

Jain proposed remote sensing and geospatial analyses and census datasets as important methods of examining farmer decision-making across India in order to draw conclusions about the relationship between climate change, groundwater depletion and agricultural production. Remote sensing is done by using satellite technology to scan crop areas for different factors such as crop area and groundwater levels. The gathered data, along with satellite photography, and historical data is processed through imagery algorithms to conduct what is known as a geospatial analysis. The census datasets Jain and her team used consist of 20,000 test wells across India.

“You have extreme [groundwater] depletion in [North India], and also South India… [India] is expected to face some of the largest impacts from warming temperatures over the coming decade,” Jain said.

Jain poses three research questions for which she provides hypotheses, quantified evidence and conclusions throughout her presentation: “What are the impacts of groundwater depletion on agricultural production?”; “Are there adaptation strategies that can reduce the negative impacts of groundwater depletion on agricultural production?”; and “What are the feedbacks between climate change, groundwater depletion, and agricultural production?”

To identify the extent to which groundwater depletion impacts agricultural production in India, Jain and her team used data on groundwater depth from 20,000 test wells across India. Data was aggregated from 2004-13.

Jain analyzed whether farmers can adapt to depleting groundwater levels by planting crops requiring less water, such as maize, than more water-intensive crops like rice. She does this by looking at whether the proportion of area planted under each crop changes per every one meter fall in groundwater depth.

“Farmers aren’t really switching crops as a way to adapt to falling water tables … you see it a little bit in the monsoon season with rice, but not so much,” Jain said.

“Groundwater depletion is associated with small yield reductions and moderate production losses for most winter crops … Even historical levels of groundwater depletion have caused crop production to decrease.” 

With regard to Jain’s second research focus, her team explored how effectively canal irrigation systems may substitute groundwater usage in depleted areas of India.

Gridded rainfall data was used to account for weather as a control in the analysis of groundwater depletion and agricultural production. Agricultural production was tracked using remote sensing measurements of winter crop areas. Groundwater depletion was measured by compiling village-level irrigation source data (a sample size of 600,000) and projections of groundwater stress.

“We used MODIS satellite data [to identify] which pixels were growing a crop or not growing a crop based on looking at vegetation phenologies [and a specific algorithm that accounted for heterogeneity],” Jain said. MODIS is short for Moderate Resolution Imaging Spectroradiometer.

Jain and her team ran cross-sectional regressions with district fixed effects to determine the effectiveness of groundwater versus canal irrigation on winter cropped areas. Factors such as soil type and average household assets at the village level were accounted for.

“Tube wells” and “dug wells” are two forms of pulling groundwater that farmers use for winter crops. Dug wells are known to pull water from shallower depths of around 8 meters and above while tube wells are more intensive and effective, and can pull groundwater from much deeper levels. 

Tube wells and canal irrigation were compared, and it was concluded that canal irrigation is associated with lower crop areas and is a less reliable irrigation method when compared to dug well irrigated crop areas in the long run. “[Canal irrigation] does really well when it rains a lot [but] groundwater villages are more stable with the amount of production they have,” Jain said.

When the data was analyzed state-by-state, tube well irrigation outperforms canals, particularly across North India. On the other hand, canal irrigation did better than or as well as groundwater irrigation in South India. 

Using data sets provided by the Indian government, Jain and her team analyzed the blocks smaller than the district-level that were categorized as overexploited or safe based on groundwater depth data. 

“If all groundwater is lost from over-exploited regions, India could lose up to 20% of winter cropped area … as an upper-bound estimate,” Jain said. The main losses are found in the North and West regions of India where most of the winter crop productions are concentrated. This area is known as the Indo-Gangetic Plains, “the equivalent of the midwestern Bread Basket,” Jain said.

The focus of Jain’s last research paper is the effect of climate change on agricultural production.

“With increasing temperatures, we expect that crop water stress would likely increase and this is because it would mean that potential evapotranspiration [would increase],” Jain said. “One way that farmers might be able to offset this increased crop water stress is to increase the amount of irrigation that they’re using [which would] likely [lead] to further groundwater declines.” 

This hypothesis was confirmed using panel fixed effects regressions and year fixed effects, where the change in groundwater levels is the dependent variable and the mean temperature is the independent variable. Potential differences in agricultural or vegetational productivity through time were controlled using the satellite-measure Leaf area index. According to Copernicus Global Land Service, “the Leaf Area Index is defined as half the total area of green elements of the canopy per unit horizontal ground area.” 

According to Jain, using projectional estimates until 2050, “we see that groundwater depletion rates could quadruple compared to current rates [of around 5 centimeters per year].” Some sensitivity analyses were conducted to see how the results would change, but the overall large magnitude of changing net groundwater levels remains.

Jain’s take home message is that groundwater depletion is a rising issue, exacerbated by warming temperatures, that is affecting India’s current crop yield and cannot be fully replaced with canal irrigation.