Impact of climate dynamics on cyclical properties of wine production in Douro region using time-frequency approach

Abstract

Abctract. In this paper we model the impact of climate dynamics on wine production temporal cycles for the period 1933 to 2013 in the Douro wine region. We identify the cyclical properties of wine production and which cycles are determined by spring temperature and soil water levels during summer. We achieve that by applying a time-frequency approach, which is based on Kalman filter regressions in the time domain. The time-varying autoregressive model can explain 79% of the variability of wine production in Douro region. We then transfer the results in the frequency domain and can show that wine production is characterized by two cycles of 5.7 and 2.5 years around the long run trend. The in-season spring temperature as well as the temperatures of two and three years ago could explain about 65% of the variability of wine production. When the soil water level in summer is incorporated, the R 2 increases to 83% and the Akaike criterion value is lower. The effects of soil water in wine production are depending on the timing. The in-season effect of an increase in soil water is negative, whilst soil water from two and three years ago have a positive effect on wine production. There is a stable but not constant link between production and the spring temperature. The temperature is responsible for two long-medium cycles of 5.8 year and 4.2 years as well as a short one of 2.4 years that began since the 80s. The soil water level can explained 60%of the 7 years cycles of wine production as well as a short one of 2.3 years cycle which has been happening since the 90s. We can recognise a shift of the relative importance away from temperature to soil water. Despite using a new an extended dataset, our results largely confirm the results of the impact of climate on the wine production in Douro region in our previous research. Modelling the impact of climate on the wine production can be an important instrument contributing for mitigation strategies facing the projected climate conditions in order to remain competitive in the market. Keywords. Climate variability, Wine production, Time-varying spectra, Kalman filter, Douro region. JEL. L52, B52, F63.