404-6 An Update on the "Abundant and Well-Distributed Rainfall" (AWDR) Parameter for the Optimization of N Rates at Sidedressing in Corn.

See more from this Division: ASA Section: Agronomic Production Systems
See more from this Session: On-Farm Research: II. Advancing Precision Ag Tools

Wednesday, November 9, 2016: 11:20 AM
Phoenix Convention Center North, Room 223

Nicolas Tremblay, Saint-Jean-sur-Richelieu Research and Development Centre, Agriculture & Agri-Food Canada, St-Jean-sur-Richelieu, QC, CANADA, Edith Fallon, Saint-Jean-sur-Richelieu Research and Development Centre, Agriculture and Agri-Food Canada, St-Jean-sur-Richelieu, QC, Canada, Rene Lacroix, Macdonald Campus, McGill University, Ste-Anne-de-Bellevue, QC, Canada, Mohamed Yacine Bouroubi, Earth Observation, Effigis GeoSolutions, Montreal, QC, Canada and Rene Audet, Agriculture and Agri-Food Canada, Quebec City, QC, Canada
Abstract:
The « Abundant and Well-Distributed Rainfall » (AWDR) parameter was proposed in Tremblay et al. (2012) as an integrator of rainfall intensity and distribution over time around N sidedressing in corn. A time window ranging from -15 days to +30 days around N sidedressing was initially proposed for maximal impact of the AWDR on corn yield response to N rate. In practice, available rainfall forecasts in Canada do not cover beyond +15 days. This -15/+15 version of AWDR was tested on an independent database of 242 site-years of N response trials in Quebec and 80 site-years in Ontario and its relationship with N rate effects was confirmed. A revisit of the most critical time window around sidedressing was made from this Quebec-Ontario database and resulted in a new estimation of -20 days to +12 days. In a real-time AWDR use framework, the difficulty resides in the calculation of AWDR not so much from the past 15 days of cumulative rainfall but also the 15 days forecasts available daily from Environment Canada servers. For Day 0 to day 5, the predictions are made using the Canadian Meteorological Markup Language (CMML) (0-24 h) and the deterministic models regional GEM (24-48 h) and global GEM (48 h-6 days). For the 6th to the 16th day, where the predictions are made by the North American Ensemble Forecast System (NAEFS), the data is supplied for each member. Some parameters control the way forecasted rain is calculated from all 40 NAEFS simulations. Based on archived daily NAEFS forecasts, a numerical optimization of AWDR calculation was conducted on 34 sites fitted with rain gauges in 2015. A total of 5923 files (1.27 GB) were retrieved in the different steps, containing 3009204 observations, and 91188 combinations of parameters and points. The retained set of parameters produced a mean AWDR deviation of 0.19 for a mean of observed AWDR of 67. The standard deviation was equal to 21.7. Predicted-observed differences varied according to dates and locations. This quantified risk on optimal N rate prediction can now be managed with a security margin of fertilizer application. The Application Programming Interface (API) for the AWDR parameter -15/+15 is currently being used by the SCAN (Soil, Crop, Atmosphere for N) decision-support system (DSS) for real-time sidedressed N rate calculations for corn in Quebec. In 2016, the AWDR estimations were lower than usual and resulted in low N rates recommendations, which correctly reflected the relatively dry season that was effectively experienced in the region.

See more from this Division: ASA Section: Agronomic Production Systems
See more from this Session: On-Farm Research: II. Advancing Precision Ag Tools