|BARBATO, LUCIA - Texas Tech University
|SESHRADI, SANTOSH - Texas Tech University
Submitted to: Meeting Abstract
Publication Type: Abstract Only
Publication Acceptance Date: 6/1/2013
Publication Date: 9/4/2013
Citation: Mahan, J.R., Payton, P.R., Barbato, L., Seshradi, S. 2013. Time surfaces of environmental and plant parameters as a tool for seasonal phenotyping: an environomics approach. Interdrought IV Conference, September 2-6, 2013, Perth, Australia. p.54.
Technical Abstract: Agricultural yield is the product of accumulated metabolic activity of the plant over the course of several months. The environment varies continuously and the plant in turn responds continuously. An improved understanding of this temporal “pathway to yield” will provide information on how and why crops respond to environmental cues. Automated monitoring of multiple variables on a 15-minute interval over a 150-day growing season can produce datasets of a million observations. The analysis of such data is currently problematic.In this project we have used a GIS-based approach to present seasonal time series data as “time surfaces” that facilitate the visualization of seasonal values of multiple variables in an interactive manner that represents a “forest before trees” analysis that allows a qualitative assessment of relatively large agriculturally-relevant data sets. We propose that time surfaces may be useful in understanding differences and similarities in crop/environment interactions across different years and geographic locations, “environomics” if you will. In this presentation we will present results of our initial efforts to visualize the interactions between cotton and its environment over multiple growing seasons in Lubbock, TX. The creation and inspection of 96 time surface visualizations of environmental variables (vapor pressure deficit, air temperature, windspeed) and plant canopy temperature provided insight into similarities and differences in plant/environment interactions. Using a time surface approach, it is relatively simple to interactively inspect datasets containing 250,000 individual observations.