Dr Chetan Deva
- Position: Research Fellow
- Areas of expertise: crop breeding; seasonal forecasting; multispectral imaging; statistics; machine learning; remote sensing; python
- Email: C.R.Deva@leeds.ac.uk
- Website: Googlescholar
Profile
I am an interdisciplinary scientist working to understand the impact of weather and climate on food systems. I use statistical learning in combination with large data sets to explore a range of topics within climate impacts. My work has focused on helping crop breeders to identify heat tolerant varieties, uncovering the genetic and environmental drivers of flowering time in plants and the development of forecasting models of crop production for early warning and early action.
I am currently a Postdoc on the CONFER project, which seeks to improve climate services for the food security sector. My role is to co-produce a model that improves upon existing capacity to forecast crop production in east Africa. I am combining large volumes of satellite observations of vegetation with advances in machine learning algorithms to work towards the achievement of this goal.
In my previous Postdoc, I worked with plant physiologists, plant breeders and rural economists to build models that can help breeders to understand stress patterns across space and time. I used hundreds of field experiments to develop a crop model for the simulation of common bean employing a combination of simple physiological equations and machine learning techniques to produce data driven algorithms for phenology and allometry.
My PhD work focused on using large data sets to understand the role of canopy cooling across spatial scales. At the plant scale, I used thousands of multispectral measurements to test the hypothesis that heat tolerant varieties of common bean are better at cooling than heat sensitive varieties. At canopy scale, I delved into the environmental physics of canopy cooling at field scale, and at regional scale, I explored the feedbacks between large areas of irrigated land and the occurrence of heatwaves.
Research interests
- Crop breeding for stress tolerance
- The genetic and environmental drivers of flowering in plants
- Integrating machine learning in modelling plant processes
- Seasonal forecasting of crop production using remote sensing and machine learning.
- Land surface – atmosphere interactions in the development of heatwaves
Qualifications
- BA Economics and Philosophy
- MA Climate and Society
- MSc Development Economics
- PhD Climate and Atmospheric Science