A collaboration with the core Nextjournal team to publish work modeling the effect of biochar on crop yields as an explorable, data driven, reproducible article.
The complex nature of biochar interactions with soils and crops as well as lack of clear understanding of the mechanics of these interactions has led to reports with conflicting interpretations, even under similar conditions. In addition, the large amount of missing data in the literature including inconsistent reporting of soil and biochar properties has made the prediction of crop yield response to biochar a very challenging modeling exercise.
We addressed this challenge by developing an extensive informatics workflow for processing and analyzing crop yield response data as well as a large spatial-scale modeling. We used probabilistic graphical models to study the relationships between soil and biochar variables and predict the probability and magnitude of crop yield response to biochar application.
Data is fed into a hybrid Bayesian Network (BN) model featuring 84 parameters. The dynamic model can propagates any new evidence through the BN and the posterior probabilities are computed. The state of the BN model is then projected onto all cultivated lands in the US based on the 2016 Cropland Data Layer using the Gridded Soil Survey Geographic soil database for the five biochar types and two application rates, i.e., 5 and 15 Mg.
An interactive map that depicts the the probability of crop yield increase following biochar application as well as the expected yield increase for every single farm around the US.
Where Should We Apply Biochar? currently includes these assets:
Biochar data, cropland data, soil data
A model to estimate the response yield of biochar applications, implemented as a Bayesian Network
Some work complete on a Shiny app intended to help users investigate predictions of biochar's effectiveness on US cropland
These existing assets and research will form the foundation of an interactive article. Our goal is to change how people engage research. To paraphrase Bret Victor, rather than "information to be consumed," the resulting notebook will be "used as an environment to think in."
This will be accomplished using Nextjournal's existing technologies and collaborative efforts with the Nextjournal team where appropriate, e.g. specialty graphics, user experience components, or back end services.
The project should roughly take 2 months to implement and begins upon receipt of the first paid installment of the award. This includes: