The five soil heat flux measurements collected over the vineyard interrow were corrected to represent a surface approximation by accounting for the heat storage in the overlying soil layer. Once G observations were corrected to the surface level, the individual half-hourly measurements of the soil heat flux were averaged to produce a representative G flux for each vineyard .ETa showed distinct daily, seasonal, and inter-annual patterns across vineyards located in different viticultural areas. Cumulative annual evapotranspiration from vineyards in California’s Central Valley was about 70% larger than the observations for the North Coast vineyards . Overall, total seasonal evapotranspiration had very similar magnitudes within vineyards of different varieties and trellising systems located in a proximal location, yet total annual ETa at the North Coast vineyards exhibit differences greater than 30%. High ETa fluxes over California’s Central Valley were expected due to larger atmospheric evaporative demands , yet the magnitude of differences found can only be explained by also accounting for irrigation and other agricultural management practices . In the Madera region, vineyards’ water use based on growing season accumulated ETa by the end of May is equivalent to the magnitude of precipitation from the previous winter season. Consequently,plastic square flower bucket atmospheric water demands during the summer months have to be satisfied through irrigation. In the summer , ETa was significantly larger in the Madera vineyards, yet the magnitude of these differences changed throughout the three years analyzed in this study .

The impact of viticultural practices on vines’ water use can only be measured directly and timely for irrigation management using ground-based sensors. Satellite remote sensing can provide daily ETa estimates at 30 m spatial resolution at most every eight days using Landsat in the best-case scenario, although fusion techniques using multiple satellite sources can extrapolate to daily estimates, but with greater uncertainty . However, such latency and uncertainty in modeled remote sensing information could be overcome by combining economical ground-based approaches such as surface-renewal along with satellite-based ETa. Differences across study sites are expected based on the variety of soils, plant density, trellising systems, plant varieties, and geographic locations . Such variability is observed across the analyzed results, yet certain characteristics patterns emerge as well. ETa fluxes while linked to meteorological conditions, they are heavily influenced by viticultural management and production goals in terms of yield and fruit quality. Given the complex interactions between soil characteristics, weather conditions, vine varieties, phenology, and management, this study highlights the importance of advancing tools to inform irrigation management based on ETa. We argue that while ETc, ETp, and ETo are also important, these variables do not provide information regarding local conditions and vineyard responses as feedback to management practices. Throughout this study, we have also found a nuanced relationship between specific viticulture management practices and vines’ water use. For instance, the timing, frequency, and magnitude of hedging and cover crop mowing led to different ETa responses throughout the assessed years and study sites.Meteorological conditions observed at the vineyards varied across different viticultural regions, and such diferences are within the magnitude of the regional meteorological conditions observed in the respective viticultural areas .

At each site, statistical analysis of mean daily temperatures and vapor pressure deficits do not suggest significant differences between years at each vineyard, which suggest that inter annual variability in overall meteorological conditions did not significantly affect atmospheric water demands throughout the analyzed growing seasons . As an exception, significant differences were observed in VPD and to a lesser degree in air temperature for the SLM site for the 2020 growing season as compared to the previous years. This difference is probably related to the re-grafting of the vines that occurred in the winter of 2020, which caused a significant reduction in λE and resulted in higher H, thus contributing to higher local VPD and air temperature. Differences in temperatures and atmospheric moisture measured at the North Coast vineyards were influenced in part by the absence of an aspirated shield in the BAR_A07 vineyard . Observed air temperature and atmospheric moisture closely resemble historical climate conditions at each study region, yet more days with extremely high temperatures and low atmospheric humidity were recorded during the analyzed period. As expected, observed air temperature and atmospheric water vapor increased and decreased, respectively, from the northern to the southern vineyards . When comparing summer months air temperature and VPD between the analyzed vineyards, we found larger differences than during the rest of the year, which we attribute to the influence of management, especially irrigation, on the local observed conditions. During the growing season, weather patterns show consistency throughout the analyzed viticultural regions . Linear regression analyses indicate significant linear relationships even between vineyards in the North Coast and Madera , yet there is a difference in magnitude of temperature and VPD as mentioned above. Precipitation decreases towards the South, yet in March and April 2020, some large precipitation events took place over the Central Valley.

However, based on relative soil water content observation at the SLM and RIP vineyards, these events do not seem to have affected long-term soil moisture conditions . Overall, greater near-surface relative soil water contents are observed in the BAR and SLM vineyards when compared to moisture conditions in the Madera fields , yet during the growing season, one of the vineyards in BAR reaches the driest conditions by mid to late summer. In contrast, the SLM vineyard remains wetter than the other vineyards throughout the year. The observations also suggest a larger range in soil moisture during the winter months, when most of the soil water relies on precipitation events. However, in August of the analyzed years, large dispersion in soil moisture was consistently observed in RIP_760 as well. This condition is related to food irrigation events during that time of the year, after harvest. This is followed by a large increase in soil moisture at RIP_720 in November which is typically food irrigated for redistributing salts and preparing for replanting of the cover crop. Flood irrigation events represent a considerable amount of the total water supply annually, which we have quantified as about to represent near a third of the total annual crop water use in the Madera vineyards. The nature of food irrigation imposes an additional level of uncertainty given the variability in infiltration capacity across the field. Unfortunately, the total amounts applied were not closely monitored by the growers. We speculate that vines might be able to tap into water stored deep in the soil when drip irrigation is not able to fully satisfy atmospheric demands early in the summer. While a dense array of soil moisture sensors up to 0.9 m deep is deployed in the Madera sites,plastic plant pot the observations seem to indicate that water infiltrates deeper in the soil which compounds with the above mentioned uncertainties. Consequently, this highlights the need for monitoring vine physiological stress indicators early in the growing season as a key information piece for irrigation management.While median and mean daily λE and H during the growing season at each vineyard remain similar in magnitude throughout the study period, the distributions of these fluxes do not present similar patterns across years or within vineyards . The Madera vineyards had a larger probability of low to negative H, which are largely compensated by high λE fluxes. Comparing the distribution of fluxes between BAR_A12 and RIP_760 provides a clear depiction of prevailing water stress conditions post-veraison in the North Coast vineyards versus near full satisfaction of atmospheric water demands in the Madera vineyards. Net radiation has a strong positive relationship between vineyards within a region, yet there were discrepancies throughout the growing season .

Observed differences were related to changes in the reflected shortwave radiation from the surface and long wave emitted terrestrial radiation. Those changes were usually related to management activities such as hedging of the vines or mowing of the cover crop. The relationship of net radiation between vineyards in different regions was not significant, and in some cases, negative to no linear trends were observed. These results highlight the importance of field-level four component radiation measurements or advanced modeling approaches to accurately represent Rn, especially when aiming to estimate other surface fluxes based on energy balance approaches.Overall cooler soil surface temperatures led to a smaller amplitude of the variable throughout the day during the growing season, which illustrates the role of soil management in land surface energy fluxes partitioning. In addition to the results highlighted in this section, a detailed list of statistical parameters comparing surface fluxes and meteorological variables at a daily frequency across sites and analyzed years is presented in Table S1. LAI is commonly regarded as a key parameter related to plant canopy processes such as radiation interception, evapotranspiration, and carbon uptake . The relationship between LAI and grapevines water use has been studied extensively , and it has been shown that trellising systems, vineyard density, pruning and hedging practices, and phenological stages, have an important effect in this relationship. Overall, we found that there was a positive relationship between LAI and ETa, yet the observed relationship shows dependency in phenological stages and particular characteristics at each vineyard. The vineyards analyzed in this study have different trellising systems, row orientations, vines densities, cover crop phenology and management . The interactions of all these factors illustrate the complexity of the relationship between these variables . Slightly closer to linear relationships between LAI and ETa were found mid to late growing season , yet only in a few cases this relationship was statistically significant. Therefore, no clear patterns between LAI and ETa were possible to distinguish from these analyses . While in the same sites and years a stronger relationship between LAI and ETa was observed , such behavior did not seem consistent throughout the study period. Large uncertainty is expected in vineyards LAI estimates based on satellite imagery, especially in vineyards with seasonal cover crops given the observed dramatic changes in relative foliage density and vertical distribution in time and space . Thus, our results might suggest that extrapolating a given relationship between LAI and ETa for calculations of a crop coefficient aiming to inform crop water demands might need to be carefully considered. In addition, in our analysis the presence of cover crops early in the season probably affects this relationship, indicating that considering this source of ET is also important .Throughout the analyzed study period and across sites, λE fluxes exhibited consistent diurnal features well coupled with solar radiation inputs . However, diurnal Hand Bowen ratios patterns suggest a potential ET enhancement due to advective conditions in June through August during the afternoon in the Madera vineyards. Under advective conditions, H is negative and behaves as an additional source of energy, which increases ET fluxes considerably when water is available. As part of the GRAPEX project, a new ongoing study is aiming to better understand and quantify the role of regional and local advection in ETa and other surface fluxes in the Madera region. Consistently low Bo throughout the growing season in the Madera vineyards suggest that ET fluxes remain near ETp . In contrast, larger estimated Boin the North Coast vineyards might indicate water stress conditions, especially at midday during July and August, which is consistent with the regulated deficit stress irrigation management targeted for these vineyards. In the SLM vineyard, Bo features different behaviors throughout the study period; moderate water stress conditions in 2018, no stress in 2019, and considerably high Bo in 2020 are related to a small canopy due to the re-grafting that took place that year, which is consistent with low observed LAI values .Regulated deficit irrigation techniques are an essential component of viticultural management for wine grape production. Fine-tuned irrigation goals in viticulture do not only focus on timely satisfying plant water demands but also manipulating water stress to accomplish production goals in terms of yield and fruit quality. In grapevines, shoot growth is extremely responsive to water stress, therefore canopy growth can be controlled by deficit irrigation. Furthermore, fruit size at harvest can be controlled by water deficit during pre-veraison, usually recommended by prescribing some degree of water deficit for a short period soon after fruit set .