Project Description:
Crop-climate modelling has a decades-long history that has contributed to our understanding of the impacts of climate change on food systems. The last 5
years or so have seen an increasing focus on nutrition security – whether or not current and future diets contain the full range of nutrients that humans
require. Hence the importance of assessing the impacts of climate change on diets, rather than a limited range of crops for which process-based models are
usually available, is now recognised. This broader range of crops includes ‘opportunity crops’ that explore potential beyond the usual staples (rice, wheat,
maize, soybean) that support nutrition security and potentially climate resilience.
Modelling the full range of opportunity crops would require many years of work. Studies have therefore employed gap-filling methods that average results
from crop-specific models to provide information about these opportunity crops. This project will improve upon this relatively simple state-of-the-art
method, and enable us to make better assessments of nutrition security.
The research will use existing data sources (crop yields, climate data and model output) to develop new methods, trial these methods with completed studies
in sub-Saharan Africa (e.g. Kenya, Zambia), and assess the difference that the new methods make. Potential methods include: a new type of crop model for
broad classes of crops such as C3 or C4 crops; machine learning; and statistical modelling. The student will be supervised by a senior postdoc and a professor
and will have the opportunity to attend weekly meetings of the climate impacts group, which include presentations and discussion as well as an informal
weekly lunch.
This project will focus on designing and evaluating options to model opportunity crops. Options include i). the development of a generic version of the
General Large Area Model for annual crops (GLAM, the in-house model developed and maintained by the research group) for C3 crops that may be
developed for comparison with gap-filling methods, and ii). a detailed analysis of the existing information on climate change impacts on fruits and vegetable
classes, supported by statistical modelling, and/or machine learning, using the latest climate change impact projections.
The student will learn computer modelling skills by being introduced to the GLAM crop model. Guidance will be provided in learning how to run the model.
The choice of methods can be made with the prior skills of the student in mind (e.g. machine learning is possible, but not essential). Crop model
developmental and growth routines can be examined to identify that which is generic across the C3 crops modelled in GLAM. Crop model parameter ranges
will be collated across crops for use in the generic model. The overall process is thus one of simplifying (i.e. generalising) from existing methods. The
important skill of judging appropriate complexity in modelling will therefore be learnt.
The model(s) will be evaluated by comparing simulated outputs to observed crop yields for a range of C3 crops. The relative skill of this method can be
assessed against the skill associated with average simulation values from crop-specific models. Once the new methods have been developed, they will be
compared to the gap-fill method, both in terms of yield projections and downstream nutrition security assessment.
In addition to the key aim of developing new methods, the student will to learn about wider issues of climate change, food systems and food security and put
the work of the placement within that wider context. This activity will take a form to be agreed with the student. Options include writing up the new results in
a format that presents the nutrition security implications of future climates and future food system scenarios; and/or presenting the work, in its wider context
and implications, to the climate impacts group. Support will be given in the development of this wider understanding.
Pre-requisites:
N/A
Supervisory Team:
Andy Challinor
Contact:
Andy Challinor: [email protected]