New Innovations for the Use of Site-Specific Data

May 5, 2014

By: Molly B Schneider

Recently several studies by AgMIP researchers have been released that highlight the development of improved methods for the use and management of site-specific data for agricultural research. Site-specific data consists of more detailed information about local environmental and economic variables that could impact production. By using site-specific data researchers will be able to create more accurate predictions of future trends in agricultural production. These articles highlight three updated techniques: production modeling, data sharing and yield assessments.

“New parsimonious simulation methods and tools to assess future food and environmental security of farm populations”, by John Antle and others, published in Philosophical Transactions of The Royal Society on Biological Sciences in 2014, outlines the development of a site-specific modeling process which will enable researchers to predict different agricultural scenarios more accurately.

Forecasting models and what-if scenarios for agriculture have been limited by the fact that traditionally, they have been developed as representations of farm behavior based on data averages. These models have not taken into account differences in agricultural situations or the tendency for individual farmers to be self-selective. While the models may be able to characterize the average impact of climatic, economic, technological, social and institutional change for a specified region, this average often fails to capture specific characterizations of different farming systems within that region.

The researchers have created an inexpensive econometric modeling process that takes into account site-specific interactions between biophysical and economic changes. Once the initial outcomes of a sample are collected directly or gathered using other observational, experimental, modeled or expert data, larger population models can be constructed. The use of site-specific data to model for a larger population allows for a more accurate scenario of agricultural responses to internal and external changes.

The researchers used the model to predict the quantifiable response of poor smallholder farmers in Kenya to changes in the production system such as crop management, fertilizer use and soil health. Researchers were able to indicate the trade-offs farmers faced when choosing whether or not to adopt the use of fertilizer for a range of prices depending on their site-specific environmental and economic conditions. This new modeling system will enable agricultural researchers to develop more accurate predictions of the response to both beneficial interventions and detrimental changes to production capacity.

“Integrated Description of Agricultural Field Experiments and Production: The ICASA Version 2.0 Data Standards”, by Jeffrey W. White and others, published in Computers and Electronics in Agriculture 2013, outlines the development of a 2.0 version of the International Consortium for Agricultural Applications (ICASA). The new version was developed to create a more inclusive and standardized data sharing system for agricultural experiments.

Agricultural experiments involve the use of countless different fields of data. There is a large number variables that need to be reported for every experiment, such as: weather conditions, soil quality, cultivation and management techniques, weeds, diseases, pests as well as crop growth. Often, scientists don’t have the means to carry out their own field experiments and must rely on the data that has been collected by others. In order to conduct their own research they require access to an inexpensive and reliable data source.

In the past, agricultural data reporting sites have lacked a standardization of data reporting, or involved a scope of variables that was too narrow or was expensive to access. In order remedy this problem, ICASA 2.0 has been developed for cataloguing the diverse range of agricultural variables in a uniform way. The new system has a larger scope of variables to choose from and a more organized method of entering, sorting and searching for data, and will be both more efficient and user-friendly. Access to the system is free, in order to encourage collaboration between data collection and use amongst agricultural scientists.

AgMIP has also developed a platform similar to ICASA Version 2.0 for data use and recording, the AgMIP Crop Experiment (ACE) database. The platform shares ICASA’s goal of providing the global community with reliable information that has been collected from thousands of international field experiments.

“Climate Adaptation Imperatives: Untapped Global Maize Yield Opportunities”, by David I. Gustafson and others, published in International Journal of Agricultural Sustainability, outlines the up-to-date assessment of the global yield gap for maize production.

The new analysis differs from ones in the past because it is based on a data-driven empirical approach rather than the simulation models that were traditionally used. This new approach allowed the researchers to get a more precise and accurate measurement of the global yield gap.

The team used data inputs that had been collected from commercial maize production trial sites and compared them to national yield levels for maize for forty-four different countries. The difference between the national yield and trial site was used to compute the yield gap for that country. Results were categorized into groups of high, medium and low yield gap countries. The research team concluded that if improvements were made to intensify maize production in low and medium yield areas to the levels of high-yield countries, the large global gap could be reduced by forty-five percent.