Understanding flowering of Sauvignon blanc in the Marlborough region, New Zealand, using high-resolution weather forecasting and the grapevine flowering véraison model

Abstract

High-resolution weather forecasting and phenological models can be combined to better understand spatial and temporal variations in the phenology of grapevine varieties. The objective of this study was to compare predictions of the time of flowering of Sauvignon blanc in the Marlborough region, New Zealand, using the Grapevine Flowering Véraison (GFV) model using temperature input data from: 1) traditional Automated Weather Stations (AWS); and 2) the Weather Research and Forecasting (WRF) model. Phenology was monitored at ten sites in 2013-14, and seven of the same sites in 2014-15, where there were corresponding AWS on site. The day of 50% flowering was determined at these sites and compared with the predicted dates simulated using the combination of the GFV model with temperature data from the AWS data and WRF models. For most sites in the central Wairau and Awatere valleys, the GFV predictions based on both temperature data sources were in agreement with observations However, there were some spatial trends in the GFV prediction bias with both temperature data sources (AWS and WRF); for example, in 2013-14 coastal and the most inland sites the predicted flowering dates were earlier than those observed. The WRF model produced differences between observations and predictions of similar magnitude to those of the AWS data and therefore provides suitable temperature input data input for phenological modelling. The agreement between AWS and WRF indicates that the observed biases are likely from the phenological model predictions, not the temperature data sources. The WRF model can therefore be used instead of AWS to generate regional maps of flowering date at 1-km resolution. This combined modelling approach can be used to integrate new phenological models, for other phenological stages, other varieties and existing or new regions, to anticipate sub-regional differences in grapevine development.