terclim by ICS banner
IVES 9 IVES Conference Series 9 Comparison of imputation methods in long and varied phenological series. Application to the Conegliano dataset, including observations from 1964 over 400 grape varieties

Comparison of imputation methods in long and varied phenological series. Application to the Conegliano dataset, including observations from 1964 over 400 grape varieties

Abstract

A large varietal collection including over 1700 varieties was maintained in Conegliano, ITA, since the 1950s. Phenological data on a subset of 400 grape varieties including wine grapes, table grapes, and raisins were acquired at bud break, flowering, veraison, and ripening since 1964. Despite the efforts in maintaining and acquiring data over such an extensive collection, the data set has varying degrees of missing cases depending on the variety and the year. This is ubiquitous in phenology datasets with significant size and length. In this work, we evaluated four state-of-the-art methods to estimate missing values in this phenological series: k-Nearest Neighbour (kNN), Multivariate Imputation by Chained Equations (mice), MissForest, and Bidirectional Recurrent Imputation for Time Series (BRITS). For each phenological stage, we evaluated the performance of the methods in two ways. 1) On the full dataset, we randomly hold-out 10% of the true values for use as a test set and repeated the process 1000 times (Monte Carlo cross-validation). 2) On a reduced and almost complete subset of varieties, we varied the percentage of missing values from 10% to 70% by random deletion. In all cases, we evaluated the performance on the original values using normalized root mean squared error. For the full dataset we also obtained performance statistics by variety and by year. MissForest provided average errors of 17% (3 days) at budbreak, 14% (4 days) at flowering, 14.5% (7 days) at veraison, and 17% (3 days) at maturity. We completed the imputations of the Conegliano dataset, one of the world’s most extensive and varied phenological time series and a steppingstone for future climate change studies in grapes. The dataset is now ready for further analysis, and a rigorous evaluation of imputation errors is included.

DOI:

Publication date: May 31, 2022

Issue: Terclim 2022

Type: Article

Authors

Luca Brillante1, Greg Jones2 and Diego Tomasi3

1Department of Viticulture & Enology, California State University, Fresno, USA
2Abacela Vineyard and Winery, Roseburg, OR, USA
3CREA-VE Research Centre for Viticulture and Enology, Conegliano, Italy

Contact the author

Keywords

phenology, climate change, time series, imputation methods, recurrent neural networks

Tags

IVES Conference Series | Terclim 2022

Citation

Related articles…

Amino nitrogen content in grapes: the impact of crop limitation

As an essential element for grapevine development and yield, nitrogen is also involved in the winemaking process and largely affects wine composition. Grape must amino nitrogen deficiency affects the alcoholic fermentation kinetics and alters the development of wine aroma precursors. It is therefore essential to control and optimize nitrogen use efficiency by the plant to guarantee suitable grape nitrogen composition at harvest. Understanding the impact of environmental conditions and cultural practices on the plant nitrogen metabolism would allow us to better orientate our technical choices with the objective of quality and sustainability (less inputs, higher efficiency). This trial focuses on the impact of crop limitation – that is a common practice in European viticulture – on nitrogen distribution in the plant and particularly on grape nitrogen composition. A wide gradient of crop load was set up in a homogeneous plot of Chasselas (Vitis vinifera) in the experimental vineyard of Agroscope, Switzerland. Dry weight and nitrogen dynamics were monitored in the roots, trunk, canopy and grapes, during two consecutive years, using a 15N-labeling method. Grape amino nitrogen content was assessed in both years, at veraison and at harvest. The close relationship between fruits and roots in the maintenance of plant nitrogen balance was highlighted. Interestingly, grape nitrogen concentration remained unchanged regardless of crop load to the detriment of the growth and nitrogen content of the roots. Meanwhile, the size and the nitrogen concentration of the canopy were not affected. Leaf gas exchange rates were reduced in response to lower yield conditions, reducing carbon and nitrogen assimilation and increasing intrinsic water use efficiency. The must amino nitrogen profiles could be discriminated as a function of crop load. These findings demonstrate the impact of plant balance on grape nitrogen composition and contribute to the improvement of predictive models and sustainable cultural practices in perennial crops.

Changes in phenolic maturity and texture characteristics of the grape berry under pre-, and post-veraison water deficit

Kékfrankos (Vitis vinifera L.) grapevines grafted on Teleki-Kober 5BB rootstock were submitted to water deficit under greenhouse conditions.

Characterization of phenolics and VOCs in wines obtained from Malbec vineyards of the Uco Valley submitted to high-altitude solar UV-B and water restriction

Characterization of phenolics and VOCs in wines obtained from Malbec vineyards of the Uco Valley submitted to high-altitude solar UV-B and water restriction

Bio-based fertilisers from fruit and vegetable residues for improving soil fertility and vine status in degraded vineyards

The H2020 RUSTICA project aims to propose, demonstrate, and implement technical solutions to convert organic residues from fruit and vegetables into high-quality novel bio-based fertilisers (BBF).

New plant protein extracts as fining agents for red wines

AIM: Quinoa (Chenopodium quinoa) is a non-allergenic pseudocereal with a high protein content