Recalibration of wheat phenology parameters

[ApsimBot]

Legacy Bug ID: 4158 Author: Bangyou Bangyou Zheng Date: 2016-05-16 13:15:33 +0000 UTC

Our validation datasets of wheat (50% heading dates or Zadoks scores near flowering ) were collected from multiple researchers across Australian wheatbelt with a wide range of environments. Previous collected datasets have been reported in Zheng et al. (2013) with new datasets included National Variety Trials from 2012 to 2013, Crop Variety Test in WA from 1985 to 2007, times of sowing trials from Ben Biddulph in 2014, sowing trials from Peter Martin in NSW.

All observations are carefully checked to remove any outliers. The massive datasets provide us an unique opportunity to test heading date prediction of the APSIM-Wheat model against the Australian wheat production system. Here we subsetted our validation datasets to 52 cultivars with 2958 observations from 202 sites. All datasets in this report are available for building a database of wheat phenlogy which could be useful to check new improvements of wheat phenlogy model.

The current parameters of wheat phenology in the released APSIM were recalibration against with the whole datasets. The RMSE is significantly improved for most of genotypes. The detail reports were presented for each genotype including the geographic and sowing distributions of observations, comparison of recalibration and released parameter values.

[ApsimBot]

Author: zhe00a Date: 2016-05-16 13:22:00 +0000 UTC

ftReCalibrationAPSIM.docx Size: 2921486

[ApsimBot]

Author: zhe00a Date: 2016-05-16 13:23:00 +0000 UTC

Zheng et al_2013_Quantification of the effects of iVRN1-i and iPpd-D1-i to predict.pdf Size: 2046392

[ApsimBot]

Author: hol353 Date: 2016-07-01 19:58:00 +0000 UTC

Wheat_new.xml Size: 112923

[ApsimBot]

Author: hol353 Date: 2016-07-01 19:59:00 +0000 UTC

I've had a look at the attached XML file from Bangyou and from a software perspective it looks OK to me.