Increased wind speed resulted in higher sensible heat losses and, therefore, berry temperatures closer to air temperature, specifically in sparse canopies . The wider row spacing in open canopies , besides providing less wind resistance, also allowed more heating of the ground and air which resulted in higher sensible heat transfer. The greater sensible heat losses in Unilateral compared to Goblet could be explained by the proximity of the Goblet clusters to the adjacent terrace slope, which has the potential for very large temperature variation that greatly affected the sensible heat fluxes. As expected, the cluster heat storage tended to be negative in the morning as the berries began to rapidly warm, and positive in the afternoon while generally cooling and releasing heat .The time series of measured and simulated berry temperature were compared graphically for east- and west-facing clusters in order to qualitatively assess model performance . Graphical results indicated good qualitative agreement between measured and modeled diurnal berry temperature variation at each site. At night time or when berries were in the shade, modeled and measured temperatures closely matched the air temperature. During intermittent periods of solar exposure, round plant pot the model was able to accurately replicate both the magnitude and duration of temperature increases over ambient.
There were a few brief periods, such as in the Unilateral vineyard, in which the timing between the measured and modeled transition from sunlit to shaded conditions lagged by about an hour. It is likely that the observed discrepancies could be explained by slight inaccuracies in the exact position of each berry and leaf. Small errors in the geometric model can translate into large errors in absorbed radiation and berry temperature during periods of sun-shade transition. The statistical error measures describing agreement between measured and modeled time series shown in Fig. 3.6 are summarized in Table 3.4. Quantitative agreement between measured and modeled berry temperature was excellent, with R2 between 0.94 and 0.97, index of agreement between 0.98 and 0.99, and NRMSE between 4.6% and 8.5%. Thus, while there could be brief periods of large error in predicted temperature at any instant when there were rapid transitions between sun and shade, their effect on daily averaged errors were relatively small.The different plant geometries and row spacing also influenced the berry heat storage and, therefore, the variability in berry temperature . Overall, including the heat storage term in the energy balance equation reduced the error of the temperature fluctuations. With no heat storage, the temperature increased or decreased too quickly, typically leading to over- or under-shooting of the berry temperature, which tended to increase model errors .The experimental data collected in this study corresponded to four field sites with different climatic and geographic conditions, and vineyard designs. The average within-canopy ambient microclimate is driven both by the local weather/climate at the site and by the canopy architecture.While it is difficult to directly compare the two Davis experiments because they were conducted during different years, the two Napa plots experienced virtually the same weather conditions and thus differences are likely to be dominated by canopy architecture.
The primary architectural differences between the Goblet and Unilateral plots were that the row spacing in Goblet was much smaller, and berries were in general much closer to the ground. This tighter row spacing and overall denser canopy led to an expected trend of lower wind speeds, more humid air, and cooler air temperatures on average. The ambient microclimate conditions have important implications for average berry temperatures because when there is minimal or no solar heating, such as at night or when the berries are in the shade, the berry temperature is nearly equal to the air temperature. Since berries spend the majority of their time with temperature near the air temperature, this period will dominate the overall average temperature of the berry. Accordingly, prior work measuring apple fruit temperature found that the long-term average fruit temperature was very close to the air temperature, although there could be large deviations from the air temperature at any instant. The differences in vineyard design not only created variability in ambient microclimate, but also introduced considerable temporal and spatial variability in berry temperature. Due to the predominantly north-south row orientation, there were often large differences between the exposed east- and west-facing clusters at any instant in time. Exposure for west-facing clusters coincided with warmer ambient afternoon temperatures, which meant that the average and maximum temperatures of west-facing clusters was typically higher than for east-facing clusters. The substantial asymmetry in temperature accumulation in north-south oriented vineyards is well-known, and has given rise to strategies based on oblique row angles aimed at achieving more even heating between both sides of the row.
Of the different variables explored, the berry temperature was also likely to be influenced by differences in wind speed created by the different vineyard geometries. The wider row spacing in open canopies provided less wind resistance and higher wind speeds that increased sensible heat losses. As observed in Unilateral compared to Goblet, the higher sensible heat losses resulted in fully-exposed berry temperatures closer to air temperature. The proximity of the clusters to the adjacent terrace slope in Goblet may have also increased berry temperatures since the ground has the potential to have high deviations in surface temperature relative to air temperature. The duration and temporal pattern of berry exposure could vary considerably depending on specifics of the berry position. Except for in the Wye vineyard, there were rarely periods in which all berries on a given side of the vine were at a similar temperature . Thus, depending on the horizontal or vertical position of the berry, and random positioning of neighboring leaves, there could be significant variability in berry temperature even on the same side of the vine.Helios simulates complex interactions between the environment and different parts of the vine from ambient weather variables that are not particularly difficult to measure, thereby making it possible to evaluate the applicability of the model at other spatial and temporal scales. Because of the spatially-explicit nature of the plant microclimate model used in this work, it was possible to resolve average differences in berry temperature due to vineyard geometry. Previous models have assumed that the berry temperature is equal to the air temperature, or used simple heuristic rules to represent the mean effect of the vines. The ability to represent berry temperature deviations from the ambient air temperature is likely important for processes that are sensitive to intermittent periods of high temperature, such as chemical composition or berry burn. Simple heuristic models are unlikely to be useful in evaluating the effects of different trellis systems, particularly systems with complex geometries such as Wye. The results of the present study showed significant differences in spatial and temporal patterns in berry temperature between trellis systems, which were well-replicated by the spatially explicit model used for prediction. The model formulation used in this work explicitly represented berry heat storage, and also compared the modeled result when berry heat storage was completely removed. As expected, removing berry heat storage resulted in much larger temporal fluctuations in response to high variability in ambient microclimate. Although inclusion of berry heat storage had a modest impact on average error metrics , it did provide a noticeable reduction in temperature variability when berries were exposed to the sun. The stabilizing effect of berry heat storage also decreased maximum berry temperatures significantly.A limitation of the proposed model is that it does not include the effect of rainfall or a wet canopy on berry temperature. However, round garden pot most quality wines are produced in regions with little average rainfall during the period of berry development. The model also did not explicitly represent the 3D variation of temperature and associated heat within the fruit as has been done in the model developed by Saudreau et al. and Saudreau et al for apple. However, validation results indicated that such detail was not necessary to achieve excellent agreement with measured temperatures, but rather an average exponential dampening of temperature fluctuations with appropriate time constant was sufficient. Another limitation of this study was that currently the model does not predict 3D spatial variations in wind speed, air temperature, or relative humidity, rather, these quantities were measured near the fruit clusters and used to drive the model.
If within-canopy microclimate was not available, it could be necessary to implement a canopy-scale energy and momentum transport model. Previous studies have shown that grape berry metabolism is sensitive to changes in both daily average temperatures and the magnitude of diurnal temperature fluctuations, therefore, advances in our understanding of the berry temperature fluctuations might help develop novel strategies to obtain the desired grape quality. The validation exercise in this work focused specifically on berry temperature from post-veraison to harvest. In future studies, the model could be modified to include latent heat fluxes and be validated to estimate berry temperature during ripening when berry evaporation may be significant and radiative properties of the berries likely differs. In addition to evaluating strategies for mitigating excessive berry temperatures, the model developed in this work could be used on the macro-scale to predict daily berry temperature fluctuations in different regions, such as hilly or mountainous areas and/or areas with arid continental weather subject to dramatic temperature fluctuations. The model could also be coupled with epidemiological and physiological models to study the effect of the spatial and temporal temperature variations on disease incidence or on physiological processes that determine grape yield and quality. Recent increases in average air temperatures and heat wave intensity can present challenges in maintaining grape productivity and quality. As a result, growers are exploring approaches to protect berries from excessive temperatures, however, they can be costly and time-consuming to evaluate experimentally and results may not be generalizable. In this work, we developed and evaluated a new 3D model that can predict metrics related to berry temperature and light interception in response to varying vineyard architecture, topography, and shade cloth density. The resulting modeling tool was applied to better understand and evaluate a range of potential vineyard design and management practices for mitigation of elevated berry temperatures in vertically-trained grapevines. Model validation showed close agreement between predicted and measured temperature dynamics, which responded appropriately to the application of shade cloth. In a simulation experiment, row spacing, row orientation, slope grade and aspect, and shade cloth density were varied in order to evaluate their effect on berry and canopy light interception, berry temperature spikes, and integrated berry heat accumulation. On flat terrain, NE-SW row orientation provided the best compromise of berry light and temperature balance between opposing vine faces while avoiding excessive berry temperatures, while N-S rows provided good daily symmetry but had risk of high afternoon berry temperatures. The efficacy of shade cloth in mitigating excessive temperatures depended strongly on all variables considered. Slopes with southern or western exposure increased imbalance and risk of high berry temperatures, which in some cases could not be well-managed by shade cloth. Overall, the modeling tool appears capable of providing quantitative guidance for vineyard design and management where excessive berry temperatures are a concern.Climate models predict that greenhouse gases will increase global average ambient temperatures by approximately 1 and 3◦C by 2030 and 2100, respectively, in addition to an increase in the frequency, duration and severity of heat waves, particularly in many wine grape growing regions. Elevated temperatures differentially affect rates of grape berry sugar accumulation and phenolic compound development, which can lead to trade-offs in harvest timing that can ultimately result in a reduction in overall grape quality. In Oakville, CA, Mart´ınez-Luscher et al. ¨ reported that elevated temperatures for grape clusters result in unbalanced wines with higher pH and lower levels of anthocyanins. Other research conducted in Murrumbidgee, Australia reported that temperatures exceeding 40◦C resulted in delayed ripening and caused berry sunburn. High ambient air temperatures can exacerbate problems created by excessive berry solar exposure due to the reduction of convective cooling. These high temperatures due to direct sunlight can result in berry cellular damage within a few minutes, while moderately high temperatures can result in injuries or death after long exposure. Grape growers have started to implement practices that modify vineyard microclimate in the short and long term to cope with elevated temperatures.