Skip to content
Frontiers
Search

Original Research

Frontiers in Horticulture ·

Volume 5 - 2026 | https://doi.org/10.3389/fhort.2026.1816162

Assessing spatial and temporal patterns of grapevine vigor and yield using satellite remote sensing

  • Daniel Kingsley Cudjoe1,2
  • Yazhou Sun1
  • Justin Scheiner3
  • Daniel Hillin4
  • Pierre Helwi5
  • Sanai Li1
  • Wenxuan Guo1,2*corresponding author
  1. 1Department of Plant and Soil Science, Texas Tech University, Lubbock, TX, United States
  2. 2Department of Soil and Crop Sciences, Texas A&M AgriLife Research, Lubbock, TX, United States
  3. 3Department of Horticultural Sciences, Texas A&M University, College Station, TX, United States
  4. 4Department of Horticultural Sciences, Texas A&M AgriLife Extension Service, Lubbock, TX, United States
  5. 5Martell & Co., Cognac, France
Article metrics0 viewsView details
PDFCiteMetrics

Abstract

Understanding spatial and temporal patterns in grapevine growth is crucial for optimizing vineyard management and improving yield and quality. This study evaluated the use of satellite remote sensing to assess spatial and temporal patterns in grapevine vigor and yield in a commercial vineyard in Terry County, Texas. Grapevine vigor was evaluated using measurements of cane length, cane diameter, and pruning weight. Yield and yield components were assessed as grape yield, clusters per vine, cluster weight, and berry weight. Using vegetation indices (VIs) derived from high-resolution PlanetScope (3 m) and moderate-resolution Sentinel-2 (10 m) satellite imagery, we assessed the spatial and temporal variability in vine vigor. The results showed significant variability in grapevine vigor, with high vigor showing greater cane length, pruning weight, and clusters per vine. The spatial autocorrelation analysis revealed significant positive clustering for cane length, pruning weight, and clusters per vine, indicating distinct spatial patterns of vine vigor within the vineyard. In contrast, cane diameter, cluster weight, grape yield, and berry weight showed weak spatial autocorrelation, suggesting minimal spatial dependence across the vineyard. Pruning weight showed the most significant relationship with grape yield, reflecting its role as an integrative indicator of vine growth. Both PlanetScope and Sentinel-2 VIs captured similar phenological patterns and spatial and temporal dynamics of grapevines. However, PlanetScope’s high spatial resolution provides greater sensitivity to subtle variability across the vineyard. Both NDVI and EVI showed comparable performance at distinguishing spatial and temporal differences among vine vigor zones, making them more effective at detecting spatial and temporal vine vigor variability during mid-to-late seasons than during early seasons. While VIs showed strong correlations with vine vigor traits during the mid-to-late seasons, their correlations with grape yield and yield components were low, indicating the need for precision management to improve vigor uniformity. These findings highlight the potential of integrating satellite-based remote sensing into precision viticulture to enhance vineyard monitoring and promote sustainable grape production.

Keywords

1 Introduction

Grapevine vigor, defined as the rate of vegetative growth and overall vine health (Winkler, 1974), plays a central role in vineyard management as it influences canopy development, fruit yield, and quality. It is shaped by multiple interacting factors, including soil properties, climatic conditions, and management practices (Pastonchi et al., 2020; Reynolds, 2022; Karn et al., 2024). Spatial variability in vigor within vineyards presents both challenges and opportunities for optimizing operations such as pruning, irrigation, and nutrient management (Karn et al., 2024). Maintaining a balance between vegetative growth and fruit production is essential for achieving consistent yield and quality (Vance, 2012; Steyn et al., 2016). However, traditional methods for assessing grapevine vigor, such as manual field measurements, are labor-intensive, time-consuming, and often inadequate for capturing spatial and temporal variability at scale (Hall et al., 2011).

Remote sensing has emerged as an effective tool for monitoring vineyard variability, enabling timely and spatial assessment of vine growth and supporting precision management (Matese and Di Gennaro, 2018; Giovos et al., 2021; Taylor et al., 2023; Trought and Bramley, 2011). Among these, satellite-based remote sensing provides consistent, large-scale observations of vine development over time (Borgogno-Mondino et al., 2018; Pastonchi et al., 2020; Tardaguila et al., 2021; Mucalo et al., 2024; Arab et al., 2021; Ni, 2024). However, satellite platforms differ in their spatial and temporal resolutions, which influence their suitability for vineyard monitoring. Moderate-resolution satellites such as Sentinel-2 (10 m spatial resolution) offer broad spatial coverage and frequent revisit times, making them well-suited for regional and temporal analyses (Sozzi et al., 2018). In contrast, high-resolution platforms such as PlanetScope (3 m spatial resolution) provide finer spatial detail, enabling improved detection of within-field variability and subtle differences in canopy structure. Previous studies have demonstrated the utility of Sentinel-2 for tracking vine phenology and identifying key growth stages (Devaux et al., 2019; Tassopoulos et al., 2021), while higher-resolution imagery has shown enhanced sensitivity to intra-vineyard variability. Therefore, careful selection of imagery is necessary depending on the scale and objectives of vineyard monitoring.

VIs derived from satellite imagery, such as the Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index (EVI), have been widely used to assess vigor, plant health, canopy structure, and biomass accumulation in vineyards (Campos et al., 2021; Wei et al., 2023; Mucalo et al., 2024). These indices provide quantitative measures of vegetation change over time, aiding early detection of stress conditions and informing management decisions. However, while these indices are effective at capturing mid-season canopy vigor, their ability to distinguish vigor zones across the season remains unexplored, particularly in vineyard systems with heterogeneous canopy structures and partial soil cover (Dobrowski et al., 2002; Barriguinha et al., 2021). The challenge is further complicated by the limitations in low spatial resolution (e.g., Sentinel-2) and spectral sensitivity inherent to each satellite platform, which may influence their capacity to detect subtle intra-vineyard differences or temporal changes in vine phenology.

Most previous studies relied on images from a single satellite, which may be limited by insufficient spatial or temporal resolution or cloud cover (Sassu et al., 2021; Araújo et al., 2023). Comparing imagery from different sources helps identify the most suitable spatial and temporal resolutions, as well as the most effective VIs, for accurately monitoring grapevine growth. Despite the increasing availability of multi-source satellite data, few studies have systematically evaluated the use of high-resolution PlanetScope and moderate-resolution Sentinel-2 imagery to monitor grapevine vigor and yield traits throughout a growing season. As such, there is a critical gap in understanding how data from different satellite platforms can be used effectively for vigor monitoring and yield prediction in vineyards. Additionally, the temporal dynamics of these indices, thus how their predictive strength varies over time, remains underexplored, especially when comparing satellites with different revisit intervals and radiometric properties. Therefore, this study aimed to assess the feasibility of using different satellite imagery to assess grapevine growth patterns and their relationship with vine vigor and yield traits. Specifically, the objectives were to: (1) quantify the spatial variability and temporal dynamics of grapevine vigor and yield components within the vineyard; (2) evaluate the performance of VIs derived from moderate-resolution Sentinel-2 and high-resolution PlanetScope imagery in capturing these patterns; and (3) assess the relationships between satellite-derived VIs and vine vigor and yield traits over time.

2 Materials and methods

2.1 Study site

This study was conducted in a 4-ha commercial vineyard near Brownfield in Terry County, Texas, in 2022 (Figure 1). According to the NRCS Soil Survey Geographic Database (SSURGO) (Soil Survey Staff, 2023), the study area consists of two distinct soil types: Amarillo fine sandy loam (AfA) and Arvana fine sandy loam (AvA). The Amarillo soil series is characterized by its deep, well-drained, loamy nature, with a notable clay and silt content that enhances its water-retention capacity. The Arvana soil series has a coarser texture, greater permeability, and lower water-retention capacity than the Amarillo soil (Soil Survey Staff, 2023). The vineyard block comprised 46 own-rooted Cinsault vine rows, each extending 288 m in length. Planted in 2015, the vineyard had a planting density of 2,157 vines per hectare, with rows spaced 3 m apart and individual vines spaced 1.5 m apart. The vineyard utilized a bilateral cordon training system.

A vineyard near Brownfield, Texas, with grapevine sampling points and vigor zones.
Figure 1. A vineyard near Brownfield, Texas, with grapevine sampling points and vigor zones.

2.2 Experimental design and grapevine sampling

To characterize the spatial patterns of vine vigor and to inform practical precision vineyard management, the vineyard was delineated into three vigor zones (low, medium, and high). The zoning was performed using k-means clustering based on a Sentinel-2 NDVI image acquired on 3 August 2021 (Figure 1). Following clustering, manual adjustments were applied to smooth minor variations (especially in the medium-vigor area) and incorporate field observations of vine growth conditions. The areas of the low, medium, and high vigor zones were 7,290 m2, 26,690 m2, and 13,550 m2, respectively. To ensure representative vigor and yield sampling, a hierarchical design was implemented with 36 locations, 12 randomly selected per zone. Each sample represented a grapevine, and the sample location was precisely determined using a differential GPS (DGPS) receiver.

At each sampling location, one grapevine was designated for measurements of vigor and yield parameters. Measurements of vigor parameters included cane length (cm), cane diameter (mm), and pruning weight (g). Yield components included clusters per vine, cluster weight (g), and berry weight (g). Grape yield (kg/ha) was calculated based on yield components and vine spacing. Manual pruning was performed in January 2023 during the dormant period, with cane measurements reflecting the vine vigor during the 2022 growing season. Individual canes were pruned from each selected grapevine. Total pruning weight was measured for each of the 36 samples.

2.3 Weather conditions and grapevine phenology

The region has a semi-arid climate, characterized by hot summers with average temperatures exceeding 30 °C and an annual precipitation of 370 mm (Kimmel et al., 2016). Weather data, including daily precipitation and maximum and minimum temperatures for the field from 1995 to 2025, were retrieved from the gridMET dataset (Abatzoglou, 2013). These data were then aggregated to compute monthly total precipitation and mean temperatures. Figure 2 compares the 2022 monthly precipitation and temperature patterns with the 30-year climate normals (1995–2025) for Brownfield, Texas. The pre-season period (January–March) was notably cold, with monthly temperatures significantly below the climatological mean, which delayed early-season grapevine development and growth. In contrast, the active growing season (April–September) experienced consistently hotter conditions, with monthly temperatures significantly higher than their long-term averages. Precipitation patterns also deviated from the long-term averages. Rainfall from January to July was substantially below the long-term averages, while August and September received substantially higher precipitation than typical for these months. These weather patterns influenced crop development and vineyard management practices.

Monthly precipitation and temperature in 2022 compared with 30-year climate normals (1995–2025) for Brownfield, Texas.
Figure 2. Monthly precipitation and temperature in 2022 compared with 30-year climate normals (1995–2025) for Brownfield, Texas.

The phenological stages of the vineyard were monitored from April to October (Table 1) using the phenological scale by Eichhorn and Lorenz (1977). Additionally, the accumulated growing degree days (GDDs) were determined at a base temperature of 10 °C.

Table 1. Phenological stages of grapevines observed in a vineyard near Brownfield, Texas, from April to October 2022.

E-L stageDescriptionDate observed
E-L 4Budbreak (leaf tips just visible)11 April 2022
E-L 125 leaves unfolded21 April 2022
E-L 1510 leaves separated25–27 April 2022
E-L 17Inflorescences clearly visible28–30 April 2022
E-L 18-19Flowering begins - 50% cap fall5 May 2022
E-L 23Full bloom10–12 May 2022
E-L 27Fruit set23 May 2022
E-L 31Berries pea-size26 May 2022
E-L 33Berries still hard and green10–15 June, 2022
E-L 35Veraison (berries begin to soften and color)15–20 July, 2022
E-L 36Berries with intermediate sugar25–28 July, 2022
E-L 37Berries not quite ripe1 August 2022
E-L 38Harvest-ripe10–15 August, 2022
E-L 39Harvest completed18 August 2022
E-L 43Leaves yellowing and senescing15 September - 10 October 2022

2.4 Management practices

Grapevine management in this vineyard followed standard practices in the Texas High Plains region. Vines were pruned annually during the study to maintain two buds per spur, six spurs per cordon, and two cordons per vine. At the end of February, initial mechanical pruning was performed to remove excess growth. This was followed by more precise hand pruning in mid-March to fine-tune the vine structure. Additionally, shoot thinning was not performed to mitigate potential hail damage, as protective hail netting was not in place during the research period. These decisions were influenced by the site’s specific environmental conditions and the need to maintain consistent vineyard management practices throughout the study. A drip irrigation system was used throughout the growing season, delivering six irrigation events spaced 10–24 days apart based on vineyard conditions. Each event applied 75 mm of water to ensure efficient moisture management. The strategy targeted two key growth phases: fruit set to veraison and veraison to harvest, which together accounted for about 70% of the vine’s total seasonal water demand, supporting fruit development and quality. To enhance vine nutrition, humic acid (12%, 7 L/ha) was applied through irrigation at bud break (E-L 4) (Eichhorn and Lorenz, 1977). Additionally, chelated zinc (12%, 3.6 kg/ha) was applied by foliar during bloom (E-L 23).

2.5 Satellite data acquisition and processing

Satellite imagery was acquired using the PlanetScope satellite constellation, operated by Planet Labs PBC (San Francisco, CA, USA), and the Sentinel-2A and Sentinel-2 B, developed by the European Space Agency (ESA) (European Space Agency, 2015) as part of the Copernicus program. PlanetScope images were downloaded from planet.com, whereas Sentinel-2 multispectral imagery was downloaded from Google Earth Engine. The PlanetScope satellite is equipped with high-resolution optical sensors that acquire multispectral imagery at 3 m spatial resolution and at high temporal resolution, enabling near-daily revisit coverage. Each PlanetScope satellite collects data in four spectral bands: blue (455–515 nm), green (500–590 nm), red (590–670 nm), and near-infrared (NIR) (780–860 nm). PlanetScope ortho scene products (Level 3B), surface reflectance products, were acquired by Dove Classic, Dove-R, and SuperDove sensors. The Sentinel-2 imagery comprises 13 spectral bands spanning the visible, NIR, and shortwave infrared (SWIR) regions. For this study, only the four 10-m resolution bands: blue, green, red, and NIR were utilized. PlanetScope and Sentinel-2 imagery for the vineyard was collected from April to October 2022, creating a time series that captured both spatial and temporal changes in grapevine development across various growth stages. Specifically, PlanetScope imagery was acquired at near-weekly to sub-weekly intervals, while Sentinel-2 imagery was obtained at approximately 5–10 day revisit intervals, depending on availability and cloud conditions.

Cloud-free satellite imagery of the vineyard was processed in Python 3.12 (Python Software Foundation, 2025). From these images, VIs, including the NDVI and EVI, were derived (Table 2). For PlanetScope, B3 (Red) and B4 (NIR) were used to derive NDVI, and B1 (Blue), B3 (Red), and B4 (NIR) were used to derive EVI. Sentinel-2A and B bands: B4 (Red) and B8 (NIR) were used to compute NDVI, while B2 (Blue), B4 (Red), and B8 (NIR) were used for EVI. These VIs were chosen for their proven effectiveness in their ability to detect key physiological and structural traits.

Table 2. Vegetation indices derived from PlanetScope and Sentinel-2 bands for the study.

Vegetation indexEquationTraitsReferences
Normalized Difference Vegetation Index (NDVI)(NIR - Red)/(NIR + Red)Leaf greenness, canopy cover, LAI, biomass, photosynthetic activity, vigor, plant health(Rouse et al., 1974)
Enhanced Vegetation Index (EVI)2.5(NIR - Red)/(NIR + 6Red-7.5*Blue + 1)Biomass, chlorophyll content, canopy structure, vigor, plant health(Jiang et al., 2008)

2.6 Data integration and statistical analysis

A 2-m polygon buffer was created around each of the 36 sample locations using ArcGIS Pro (Version 3.6) (Esri, 2025). This buffer size was used to ensure that each ground measurement corresponded to the area occupied by an individual grapevine. VIs were extracted for each buffered polygon. The dataset, including yield and yield components, as well as pruning parameters such as cane length, cane diameter, and pruning weight, was incorporated into a shapefile and a CSV file to support subsequent analyses. The kriging interpolation method in R software (R Core Team, 2012) was used to map the spatial distribution of vigor and yield parameters within the vineyard.

Spatial autocorrelation of the vigor and yield parameters was conducted using Moran’s I. Statistical analysis was performed using R (R Core Team, 2012). Descriptive statistics, including means, standard deviations (SD), and coefficient of variation (CV), were computed for cane length, cane diameter, pruning weight, berry weight, cluster weight, clusters per vine, and grape yield from each vigor zone comprising 12 samples. This approach was adopted to better characterize within-zone variability and to highlight differences in vine performance across the delineated vigor classes. ANOVA was performed using the aov function to assess differences in vine parameters across 36 sampling locations within three vigor zones. Pearson’s correlation was performed across 36 samples to examine the relationships between VIs and measured grapevine parameters.

3 Results

3.1 Assessing the variability of grapevine traits

Table 3 presents summary statistics for grapevine traits and grape yield across different vigor zones in the 2022 season. Clear trends were observed in vine vigor traits, where cane length and pruning weight showed higher mean values of 87.1 cm and 1295.8 g, respectively, in the high-vigor zone. The medium-vigor zone recorded a slightly higher mean cane diameter (7.5 mm) than the other zones, indicating moderate vigor. Within-zone variability was moderately high across most traits, particularly for cane length (CV = 15.5%) and cane diameter (CV = 13.3%) in the high-vigor zone. These elevated CVs highlight localized differences in vine growth, likely driven by soil heterogeneity or microclimatic variation. In contrast, pruning weight showed the lowest variability (CV = 4.3%) in the medium-vigor zone, reflecting more uniform vine performance. Among yield traits, grape yield and clusters per vine recorded the highest means in the high-vigor zone (21417.5 kg/ha and 14.6 clusters per vine, respectively), consistent with the general association between vegetative vigor and yield potential. However, cluster weight was higher in the low-vigor zone (137.8 g). In comparison, berry weight recorded similar values across vigor zones (2.1-2.2 g). These patterns of vine vigor and berry attributes suggest that excessive canopy growth can limit berry size and quality through shading and competition for assimilates.

Table 3. Summary statistics of grapevine traits under low, medium, and high vigor zones during the 2022 season.

TraitLowMediumHigh
MeanSDCV%MeanSDCV%MeanSDCV%
(n=12)(n=12)(n=12)
Cane length (cm)69.97.610.869.97.610.887.113.515.5
Cane diameter (mm)6.90.913.37.50.911.77.11.013.3
Pruning weight (g)635.065.510.3842.136.04.31295.8103.38.0
Cluster weight (g)137.841.730.3116.936.231.0137.151.637.6
Grape yield (kg/ha)12156.76680.555.016538.09515.757.521417.521477.3100.3
Berry weight (g)2.20.311.72.20.313.42.10.523.1
Clusters per vine8.83.337.314.55.739.114.67.551.7

The overall variability among yield traits was considerable. Grape yield showed the highest relative variability, followed by clusters per vine across all vigor zones, indicating strong spatial heterogeneity in fruit production within the vineyard. Berry weight showed the lowest variability (CV = 11.7%) in the low vigor zone, suggesting it is the least influenced by spatial factors. Collectively, these variations demonstrate that vineyard heterogeneity strongly influences vegetative and reproductive traits, underscoring the importance of site-specific management.

Analysis of variance (Table 4) revealed significant differences in cane length (p < 0.001) and pruning weight (p < 0.0001) across vigor zones. No significant difference was observed for cane diameter (p = 0.36), suggesting that this trait is less sensitive to distinguishing vigor levels. For yield-related traits, only clusters per vine differed significantly among zones (p < 0.05), with higher values observed in high-vigor zones. The lack of significant differences for grape yield, cluster weight, and berry weight (p > 0.05) may be attributed to the compensatory effects among yield components, where fewer but larger clusters, or smaller clusters with more berries, balanced total yield. The combined descriptive and ANOVA results indicate that vine vigor exerts a stronger influence on vegetative growth than on yield or berry morphology. Therefore, while vigor enhances canopy development, it does not always translate directly into higher yields or berry size in uniformly managed vineyards.

Table 4. Summary of ANOVA for the effects of vigor zones on vine traits for the 2022 season.

TraitSourceDFSum SqMean SqF-valuePr (>F)
Cane length (cm)Zone223541177.211.95<0.001
Residual33325198.5
Cane diameter (mm)Zone21.7620.8811.0530.36
Residual3327.610.837
Pruning weight (g)Zone22.74 × 1061.37 × 106252.8<0.0001
Residual331787375416
Cluster weight (g)Zone2337916890.8870.421
Residual33628341904
Grape yield (kg/ha)Zone25.15× 1082.58 × 1081.2950.287
Residual336.56 × 1091.99 × 108
Berry weight (g)Zone20.0320.01580.120.887
Residual334.3480.1318
Clusters per vineZone2260.7130.43.919<0.05
Residual331097.633.3

Statistically significant results (p < 0.05) are shown in bold.

3.2 Spatial variability of grapevine vigor and yield

Table 5 shows the spatial autocorrelation (Moran’s I) of the vine traits. Vine traits, including cane length (Moran’s I = 0.137, p = 0.0276), pruning weight (Moran’s I = 0.685, p < 0.001), and clusters per vine (Moran’s I = 0.232, p = 0.0011), showed significant positive spatial autocorrelation, reflecting notable spatial variability in vine vigor across the vineyard. This spatial structure suggested that vegetative growth was influenced by underlying environmental factors, such as soil fertility, moisture availability, and topography, as well as the vigor zone. Consequently, distinct zones of vine vigor were evident within the field, which could inform site-specific management to improve uniformity and resource efficiency. In contrast, grape yield showed weak spatial autocorrelation (Moran’s I = 0.084, p = 0.0521), which was not statistically significant, indicating little spatial dependence. Reproductive traits, including berry weight, cluster weight, and clusters per vine, also showed weak or random spatial patterns, suggesting they were more influenced by microenvironmental or physiological factors than by spatial position.

Table 5. Spatial autocorrelation (Moran’s I) of vine traits.

TraitMoran’s IP-value
Cane length (cm)0.1370.0276
Cane diameter (mm)–0.0140.4335
Pruning weight (g)0.685<0.001
Cluster weight (g)–0.0880.7597
Grape yield (kg/ha)0.0840.0521
Berry weight (g)–0.1050.8119
Clusters per vine0.2320.0011

Significant spatial autocorrelations at p < 0.05 are in bold.

Maps derived from the interpolated vine measurements revealed different spatial patterns in vine vigor parameters across the vineyard (Figure 3). Cane length was higher (100–115 cm), concentrated in the northeastern region, extending slightly toward the southeast. Lower cane length values (50–60 cm) were observed along the northwestern and northeastern borders, corresponding to areas of reduced vine growth. This pattern suggests underlying variation in soil moisture or nutrient availability, as these parameters strongly influence vegetative development. Similarly, cane diameter displayed a heterogeneous spatial pattern, with higher values (8–9 mm) distributed in the eastern, southeastern, and central portions of the vineyard. Areas with reduced cane diameter (5–6 mm) were primarily located in the northern and western regions, indicating relatively lower vine vigor. Pruning weight followed a spatial distribution similar to that of cane length, with the highest biomass accumulation (1400–1450 g) occurring in the northeastern to southeastern regions, coinciding with high-vigor zones established by Karn et al. (2024), while the lowest values (600–700 g) were found in the southwestern section, reflecting the established low-vigor zone. These spatial patterns in vigor traits suggest strong local autocorrelation and environmental influence, likely related to vineyard topography, soil depth, or micro-irrigation efficiency (Karn et al., 2024).

Spatial distribution of cane length **(A)**, cane diameter **(B)**, and pruning weight **(C)** in a vineyard near Brownfield, Texas.
Figure 3. Spatial distribution of cane length (A), cane diameter (B), and pruning weight (C) in a vineyard near Brownfield, Texas.

Yield and yield components also demonstrated distinct spatial variability patterns (Figure 4). Grape yield was high (80,000–84,000 kg/ha) and concentrated in the northeastern and southeastern regions, corresponding to high-vigor zones. Conversely, lower yields (5,000–10,000 kg/ha) were observed in the western, southwestern, and central parts of the vineyard, which aligned with areas of low to moderate vigor. This spatial correspondence between vigor and yield highlights the influence of vegetative growth on fruit productivity. Among the yield components, cluster weight was high (190–250 g) in the northern and northeastern regions, overlapping both high and low vigor areas. This suggests that factors beyond vegetative vigor, such as microclimatic differences or fruit set efficiency, may also have contributed to yield variability. Berry weight followed a more scattered pattern, with localized high values (2.8–3.0 g) in the southeastern, southern, and western regions, while lower values (1–1.5 g) appeared in similar areas, indicating fine-scale variation possibly driven by differential water availability or sunlight exposure. Clusters per vine displayed the most distinct spatial variability, with higher counts (30 clusters) concentrated in the northeastern section and lower counts (4–8 clusters) in the southwestern region.

Spatial distribution of **(A)** clusters per vine, **(B)** cluster weight, **(C)** berry weight, and **(D)** grape yield in a vineyard near Brownfield, Texas.
Figure 4. Spatial distribution of (A) clusters per vine, (B) cluster weight, (C) berry weight, and (D) grape yield in a vineyard near Brownfield, Texas.

3.3 Temporal and spatial patterns of satellite imagery vegetation indices

Figure 5 shows the NDVI maps for PlanetScope on 5 May, 6 June, 7 July, 5 August, and 4 September, and for Sentinel-2 on 5 May, 9 June, 4 July, 3 August, and 7 September. The high-resolution PlanetScope NDVI showed clear temporal dynamics and fine-scale spatial variability across the growing season. In early May, NDVI values were relatively low and spatially fragmented, indicating limited canopy development but soil pixels. By early June, NDVI increased across the field, with clear emergence of spatial variations and localized zones of high values. This pattern intensified in July and August, when NDVI reached peak values and exhibited strong spatial contrasts, including distinct patches and gradients reflecting subtle variations in canopy density. By early September, NDVI declined across much of the field, consistent with seasonal transition and reduced canopy activity, although localized areas of relatively higher values persisted. In comparison, Sentinel-2 NDVI exhibited similar temporal trends but with smoother, more generalized spatial patterns. NDVI values increased from May through July, peaked during mid-season, and declined toward September, reflecting the seasonal patterns observed in PlanetScope-based NDVI. However, spatial variability was less pronounced due to the coarser resolution of Sentinel-2, which produces larger pixel footprints that blend canopy and soil contributions, reducing fine-scale contrast. Therefore, while Sentinel-2 effectively captured overall temporal dynamics and spatial gradients, it tended to mask fine-scale variability that was more clearly resolved in higher-resolution PlanetScope imagery.

Spatial distribution of NDVI values obtained from PlanetScope and Sentinel-2 from May to September 2022, within a vineyard near Brownfield, Texas.
Figure 5. Spatial distribution of NDVI values obtained from PlanetScope and Sentinel-2 from May to September 2022, within a vineyard near Brownfield, Texas.

NDVI and EVI derived from PlanetScope and Sentinel-2 showed clear temporal and spatial patterns in connection with grapevine phenology (E-L 3 to E-L 43) and accumulated GDD from April to October 2022 (Figure 6). Accumulated GDD increased steadily from April (<100 °C) to October (2600 °C), closely reflecting the seasonal pattern of grapevine development. The rapid increase in GDD from May to July coincides with the period of the strongest increase in VI, reflecting active vegetative growth as thermal accumulation rises. The stabilization of GDD accumulation after August paralleled the plateau and subsequent decline in VI values associated with harvest, regrowth, and senescence. The dynamics of VIs reflected the spatial and temporal vine vigor within the growing season.

Temporal and spatial patterns of NDVI and EVI derived from PlanetScope and Sentinel-2 imagery at different vigor zones and the accumulated GDD from April to October 2022, corresponding to key vineyard phenological stages for a vineyard near Brownfield, Texas.
Figure 6. Temporal and spatial patterns of NDVI and EVI derived from PlanetScope and Sentinel-2 imagery at different vigor zones and the accumulated GDD from April to October 2022, corresponding to key vineyard phenological stages for a vineyard near Brownfield, Texas.

During early growth (April to early May; E-L 3-15), both indices were low and showed a relatively flat trend, with minimal separation among vigor zones due to limited canopy development. NDVI values during this period ranged from approximately 0.15–0.19 for PlanetScope and 0.15–0.17 for Sentinel-2, while EVI values were lower and more compressed, typically 0.10–0.13 for both sensors. A separation among high, medium, and low-vigor zones emerged from early May (E-L 18) when flowering began. At this growth stage, a sharp increase in PlanetScope NDVI values was observed compared to the smooth rise from Sentinel-2. During the same period, EVI values from both satellites increased, but with much lower slopes.

By the end of May and mid-June (E-L 31-33), at Berries pea-size, PlanetScope NDVI reached 0.32–0.33 in the high-vigor zone, compared to 0.31–0.32 (medium-vigor zone) and 0.30–0.31 (low-vigor zone), demonstrating more apparent separation and strong sensitivity to canopy vigor. Sentinel-2 NDVI showed a similar pattern, but at slightly lower magnitudes. EVI increased more modestly over the same period, with PlanetScope EVI around 0.24 in the high-vigor zone, 0.23 in the medium-vigor zone, and 0.22 in the low-vigor zone. Sentinel-2 EVI also showed separation, but with a narrower spread (high = 0.21, medium = 0.21, low = 0.18). All indices declined from late June to early July (E-L 33) when berries were still hard and green. A severe hail-producing thunderstorm passed through the region on June 3, 2022, causing substantial grapevine damage. This impact was clearly captured by satellite imagery, with both NDVI and EVI showing pronounced declines in early June, particularly EVI.

VI values reached their seasonal peak between July and early August (E-L 35-37). The high-vigor zone consistently recorded the highest values compared with the medium and low-vigor zones, as captured by both satellites. NDVI peaks were consistently higher than EVI but followed the same spatial pattern. For example, PlanetScope NDVI peaked near 0.38 in the high-vigor zone, 0.32 in the medium-vigor zone, and 0.29 in the low-vigor zone. PlanetScope EVI reached 0.24 in the high vigor zone, 0.19 in the medium vigor zone, and 0.16 in the low vigor zone, reflecting a clearer separation between vigor zones during the peak season. A similar pattern was observed for Sentinel-2 NDVI and EVI. This period corresponded to veraison, when physiological activity and canopy density reached their maximum, and GDD had accumulated to approximately 1,800–2,200 °C.

Following the peak in VI values, both NDVI and EVI declined in early to mid-August, corresponding to pruning before harvest and the removal of vegetative material during harvest (E-L 37–39). VI values increased again during late August and September (E-L 39-41). For instance, in the high-vigor zone, PlanetScope NDVI rose from 0.39 to 0.41, and Sentinel-2 EVI increased from 0.23 to 0.28, indicating renewed leaf growth, leaf area expansion, and recovery of canopy greenness. A similar trend was observed for Sentinel-2 NDVI and PlanetScope EVI. This regrowth was driven by substantial late-season rainfall, with both August and September receiving more than 100 mm of precipitation (Figure 2). A final decline in VI values occurred in October (E-L 43) as vines entered senescence, reflecting reduced leaf chlorophyll content and the progression toward dormancy. Throughout the season, NDVI values were generally higher than EVI values, but both indices effectively tracked phenological development and maintained clear separation among vigor zones, particularly in the mid- to late-season.

3.4 Relationship between satellite vegetation indices and vine traits

The correlation between VIs and vine vigor, yield, and yield components varied across growth stages, reflecting vine canopy development, spectral sensitivity, and image spatial resolution (Figure 7). During the early growth phase, correlations between VIs and vine parameters were generally weak and inconsistent. For PlanetScope, both NDVI and EVI showed low positive correlations with cane length and pruning weight (r < 0.35). Similarly, Sentinel-2 NDVI and EVI showed low correlations on 5 May, which became negatively significant by 15 May, particularly for cane length (r = –0.49 to –0.51, p < 0.05) and pruning weight (r = –0.60 to –0.62, p < 0.05). These negative relationships likely resulted from heterogeneous leaf emergence and the strong influence of soil reflectance. At this stage, correlations with yield and yield components were also weak to moderate. PlanetScope NDVI and EVI showed moderate correlations with grape yield (r = 0.32–0.33) and clusters per vine (r = 0.25–0.42), whereas Sentinel-2 indices showed moderate correlations with berry weight (r = 0.34–0.35). Therefore, early-season VIs had limited predictive value, particularly PlanetScope, for vigor and yield before full canopy establishment.

Pearson correlation coefficients between NDVI and EVI derived from PlanetScope and Sentinel-2 imagery and vineyard vigor, yield, and yield components from May to October 2022. **, * indicate significant levels at p < 0.01, 0.05, respectively.
Figure 7. Pearson correlation coefficients between NDVI and EVI derived from PlanetScope and Sentinel-2 imagery and vineyard vigor, yield, and yield components from May to October 2022. **, * indicate significant levels at p < 0.01, 0.05, respectively.

Correlations between VIs and vigor parameters peaked during veraison (July 14 to August 5) when the canopy reached its maximum density and photosynthetic activity. Both NDVI and EVI from PlanetScope and Sentinel-2 showed significant positive correlations with vigor parameters, particularly pruning weight (r = 0.80–0.86, p < 0.01) and cane length (r = 0.70–0.72, p < 0.01). These relationships confirm that midseason VIs effectively captured vine vigor and biomass accumulation. PlanetScope NDVI and EVI (r = 0.86, p < 0.01) showed slightly higher correlations than Sentinel-2 NDVI (r = 0.80–0.84, p < 0.01) and EVI (r = 0.80–0.85, p < 0.01), reflecting the advantage of finer spatial resolution (3 m) in capturing within-row variability and minimizing soil interference. Conversely, cane diameter showed weak relationships (r < 0.25) for both VIs. Correlations with yield and yield components during this period were moderate, with clusters per vine and grape yield showing r values between 0.25 and 0.45. These moderate associations suggest that vine vigor does not fully explain fruit-level variability.

During the late season (September to October), when leaves began to yellow and senescence, correlations remained positive but declined slightly relative to midseason. Both PlanetScope and Sentinel-2 indices showed good correlations with cane length (r = 0.60–0.68, p < 0.01) and strong correlations with pruning weight (r = 0.75–0.83, p < 0.01), indicating that residual canopy greenness and structure continue to reflect vine vigor as physiological activity declines. However, relationships with cane diameter and fruit traits remained weak (r < 0.40). PlanetScope-derived indices consistently showed stronger and more stable correlations than Sentinel-2 indices. Sentinel-2, however, demonstrated similar trends with slightly lower correlation magnitudes.

3.5 Correlation between vine vigor and yield traits

Pearson correlation analysis showed clear patterns of intra- and interrelationships between grapevine vigor and yield components within the vineyard during the 2022 season (Table 6). Among vigor traits, strong intra-correlations were observed, with cane length positively associated with pruning weight (r = 0.68, p < 0.001) and moderately with cane diameter (r = 0.35, p < 0.05). Within yield components, strong correlations existed between grape yield and clusters per vine (r = 0.88, p < 0.001) and cluster weight (r = 0.78, p < 0.001). Cluster weight was also positively correlated with clusters per vine (r = 0.51, p < 0.01). The correlation between vigor and yield traits was more variable. Pruning weight showed a low but significant positive correlation with grape yield (r = 0.35, p < 0.05) and clusters per vine (r = 0.39, p < 0.05), indicating that vines with greater vegetative growth tended to produce higher yields. In contrast, cane length (r = 0.15, p > 0.05) and cane diameter (r = –0.05, p > 0.05) were not significantly associated with grape yield. Berry weight showed no significant correlation with either vigor or yield traits (e.g., r = –0.16 to –0.08, p > 0.05), indicating that berry size is largely independent of vine vigor. Overall, yield is primarily driven by yield components, particularly clusters per vine and cluster weight. The results suggest that pruning weight is the most consistent vigor trait, reflecting its role as an integrative measure of vine growth. The low correlation between yield and vine vigor indicates the potential of site-specific management strategies to optimize yield and fruit quality, particularly in zones with excessive vegetative growth.

Table 6. Correlation between grapevine vigor and yield traits within the vineyard in Terry County, Texas, for the 2022 season.

TraitCane length (cm)Cane diameter (mm)Pruning weight (g)Cluster weight (g)Grape yield (kg/ha)Berry weight (g)Clusters per vine
Cane length (cm)0.35*0.68***0.050.15-0.160.19
Cane diameter (mm)0.35*0.13-0.23-0.05-0.190.07
Pruning weight (g)0.68***0.130.100.35*-0.120.39*
Cluster weight (g)0.05-0.230.100.78***-0.080.51**
Grape yield (kg/ha)0.15-0.050.35*0.78***-0.090.88***
Berry weight (g)-0.16-0.19-0.12-0.08-0.09-0.10
Clusters per vine0.190.070.39*0.51**0.88***-0.10

*, , * indicate significant levels at p < 0.001, 0.01, 0.05, respectively.

4 Discussion

This study aimed to assess the potential of satellite-based remote sensing to quantify the spatial and temporal variability in grapevine vigor and yield components, and to assess relationships between VIs and vine traits. To address this, a zone-based experimental design (low, medium, and high vigor) with 36 georeferenced sampling locations was combined with multi-temporal imagery from high-resolution PlanetScope and moderate-resolution Sentinel-2 across key phenological stages. The findings showed significant spatial variability in cane length, pruning weight, and the number of clusters per vine. High-vigor areas showed increased cane length, pruning weight, and yield, aligning with a previous study that linked increased vegetative growth with higher productivity (Bonilla et al., 2015). Cluster weight was slightly greater in low-vigor zones, which may indicate lower competition among clusters that resulted in larger cluster development, a trend noted in a previous study (Gatti et al., 2022). The lower variation in yield components, such as berry weight, suggests that berry development may be influenced by microclimatic factors other than vigor alone (Smart, 1985; Filippetti et al., 2013). These results support previous work suggesting that vegetative vigor, soil properties, microclimate, and water availability drive within-field variability (Karn et al., 2024). The presence of variability in the vineyard suggests that site-specific management practices may be needed to optimize vineyard productivity and sustainability.

The correlation patterns observed in this study highlight the complex relationships between vine vigor and yield in semi-arid vineyard systems. Pruning weight emerged as the most informative vigor metric, highlighting its usefulness as an integrative indicator of vegetative growth and productive capacity. In contrast, structural traits such as cane length and cane diameter showed limited or no association with grape yield, suggesting they are weaker predictors of yield under the conditions of this study. The strong correlations between grape yield, cluster weight, and clusters per vine further emphasize the role of crop load in driving yield variability.

Satellite-derived VIs showed varying degrees of correlation with the measured vine parameters, indicating their potential for vineyard monitoring but also their limitations under specific conditions and times of the season. The correlation analysis showed that NDVI and EVI were strongly associated with vigor traits in the mid-to-late season but showed weaker correlations with yield and yield components. This is consistent with a study indicating that while NDVI effectively tracks canopy density, it does not always correlate well with yield due to confounding factors such as berry set and cluster architecture (Hall et al., 2011). Low correlations between VIs and grape yield suggest the potential for site-specific pruning to reduce vigor in high-vigor areas identified by satellite imagery. This approach can improve fruit quality while maintaining yield and extending the productive lifespan of the vineyard.

The temporal and spatial patterns of NDVI and EVI relative to accumulated GDD highlight their responses to grapevine phenology, vine growth, weather conditions (such as hail and late-season rainfall), and management practices during the 2022 growing season. Although both PlanetScope and Sentinel-2 captured similar temporal and spatial patterns and distinguished different vigor zones, subtle differences in magnitude and variability indicate the influence of sensor characteristics. PlanetScope’s finer spatial resolution likely enhanced sensitivity to within-zone canopy variability, leading to higher NDVI values, particularly in high-vigor zones where dense foliage dominates the pixel footprint. Sentinel-2, with its coarser resolution and broader spectral bands, produced smoother seasonal trajectories that may better represent zone-scale averages but slightly dampen extreme values. Differences in band placement, radiometric calibration, and atmospheric correction further shift absolute NDVI values upward, particularly in PlanetScope, even when both sensors capture similar seasonal and spatial trends. On the other hand, EVI includes a blue band correction and soil/canopy background adjustment, along with scaling coefficients that reduce the magnitude of the index. NDVI and EVI peaked mid-season, reflecting maximum canopy cover, consistent with previous findings linking these indices to peak vegetation activity (Lu et al., 2015; Di Gennaro et al., 2019; Hu et al., 2021). Our results suggest that satellite images are more effective at capturing variability in vine vigor during mid-to-late seasons than during early seasons. These findings demonstrate the complementary roles of NDVI and EVI in precision viticulture, with mid-season peaks serving as reliable indicators for canopy maturity and vegetation health. Our results further suggest that PlanetScope and Sentinel-2 imagery could be used alternatively for vineyard assessment, depending on their availability. The application of satellite imagery in vineyards faces challenges due to the discontinuous structure of grapevine canopies, their partial ground coverage, and the substantial influence of background elements and shadows on the reflectance signal. Di Gennaro et al. (2019) previously identified mixed-pixel issues caused by the row-based layout of vineyards during monitoring of a Montepulciano vineyard with a tendone trellis system. The spatial resolution of Sentinel-2 (10 m) may not fully capture within-row variability, potentially reducing its effectiveness in vineyard monitoring compared to higher-resolution platforms such as UAV-based imagery (Matese and Di Gennaro, 2018; Mucalo et al., 2024).

From a practical perspective, the findings highlight the importance of integrating remote sensing data to optimize vineyard management practices. This study also contributes to the growing body of research on remote sensing applications in viticulture, emphasizing the need for multi-spatiotemporal analyses to improve the accuracy of yield predictions. One limitation of this study is its reliance on a single growing season, which may not capture interannual variability driven by climatic fluctuations. Interannual climatic variability, particularly differences in precipitation, may influence the magnitude of vine responses and VI–trait relationships. However, the overall patterns observed in this study, including the spatial structure of vigor, the temporal behavior of VIs across phenological stages, and the relative performance of the satellite platforms, are driven by fundamental canopy development processes and are therefore expected to be broadly consistent across seasons. While absolute values and correlation strengths of variables assessed in this study may vary under different precipitation regimes, the general trends and methodological insights remain transferable, and we emphasize the need for multi-year validation in future studies. Future research should explore multi-year studies to assess interannual variability and its impact on vigor-yield relationships.

The integration of additional remote sensing technologies, such as UAV multispectral and hyperspectral imaging, LiDAR, and other VIs, could enhance the characterization of canopy structure and vine vigor (Di Gennaro et al., 2019; Khaliq et al., 2019; Sozzi et al., 2020; Araújo et al., 2023; Ferro et al., 2023; Matese et al., 2022). UAV imagery offers finer spatial resolution and greater flexibility in data acquisition, which can capture detailed within-zone variability, while satellite data provides consistent, large-scale temporal coverage. Microwaves can provide valuable information about plant canopy structure due to their ability to penetrate the canopy. When combined with visible and near-infrared (NIR) data, microwave remote sensing enables a more comprehensive characterization of crop conditions by integrating structural and spectral information, thereby improving monitoring and analysis (Guo et al., 2023). Integrating soil brightness, ground cover, and thermal imagery enables robust assessment of water stress across a field (Torrion et al., 2014), helping to characterize the spatial and temporal dynamics of grapevine conditions throughout the growing season for precision management. Furthermore, combining remote sensing data with soil moisture and microclimate sensors could improve monitoring and interventions. Further study into machine learning models for fusing multi-source data may also enhance predictive capabilities in vineyard monitoring.

5 Conclusions

This study demonstrated the potential of satellite remote sensing in monitoring grapevine phenology, vigor, and yield. The results showed significant spatial variability in vine parameters, including cane length, pruning weight, and clusters per vine. Pruning weight was the most significant vigor trait associated with grape yield, reflecting its role as an integrative measure of vine growth. By leveraging VIs derived from PlanetScope and Sentinel-2 imagery, this study identified significant spatial and temporal patterns in grapevine phenology and vigor. Both PlanetScope and Sentinel-2 indices showed comparable phenological patterns and captured the spatial and temporal dynamics of grapevines. NDVI and EVI showed similar performance at distinguishing spatial and temporal differences among vine vigor zones, making them more effective at detecting vine vigor variability during mid-to-late seasons than during early seasons. Additionally, both indices correlated strongly with vine vigor traits but showed low correlation with yield.

The weak relationships between VIs and yield suggest that site-specific pruning and application of inputs, such as nutrients and water, especially in the high-vigor zones indicated by the satellite imagery, could help manage excessive vine growth. This strategy has the potential to enhance fruit quality, sustain yield, and support the long-term productivity of the grapevines. Our findings demonstrate that satellite images have the potential to capture vine responses to hail or other weather events (such as late-season rainfall) and to pruning or harvesting management. Future research should examine how combining additional VIs from multiple sensor platforms with machine-learning approaches can further strengthen prediction accuracy for vineyard monitoring.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Ioannis Charalampopoulos, Agricultural University of Athens, Greece

Reviewed by: Ron Shmuleviz, University of Verona, Italy

Mehdi Shabanian, University of Guelph, Canada

Statements

Conflict of interest

Author PH was employed by the company Martell & Co.

The remaining authors declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

The author(s) declared that financial support was received for this work and/or its publication. This study was funded by the Texas Department of Agriculture (Award number 21ARED13) and USDA NIFA (Award: 9898).

Data availability

The original contributions presented in the study are included in the article/supplementary material. Further inquiries can be directed to the corresponding author.

Author contributions

DC: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Software, Validation, Visualization, Writing – original draft, Writing – review & editing. YS: Data curation, Formal Analysis, Investigation, Methodology, Software, Visualization, Writing – original draft, Writing – review & editing. JS: Conceptualization, Funding acquisition, Investigation, Methodology, Validation, Writing – review & editing. DH: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Writing – review & editing. PH: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Software, Visualization, Writing – review & editing. SL: Conceptualization, Data curation, Formal Analysis, Investigation, Methodology, Software, Visualization, Writing – review & editing. WG: Conceptualization, Data curation, Formal Analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.

Acknowledgements

We appreciate Andy and Dusty Timmons and Lost Draw Vineyards for providing the study site and their valuable assistance.

References

  1. Abatzoglou J. T. (2013). Development of gridded surface meteorological data for ecological applications and modelling. Int. J. Climatol. 33, 121 – 131. doi:. PMID: doi:10.1002/joc.3413
  2. Arab S. T., Noguchi R., Matsushita S., Ahamed T. (2021). Prediction of grape yields from time-series vegetation indices using satellite remote sensing and a machine-learning approach. Remote Sens. Appl.: Soc. Environ. 22, 100485. doi:. PMID: doi:10.1016/j.rsase.2021.100485
  3. Araújo J., Norberto M., Pinheiro P., Mano C., Campos J., Pimenta V.. (2023). “ Satellite and UAV-based anomaly detection in vineyards,” in BIO web of conferences, vol. 68. (Les Ulis, France: EDP Sciences), 01022. doi: doi:10.1051/bioconf/20236801022
  4. Barriguinha A., de Castro Neto M., Gil A. (2021). Vineyard yield estimation, prediction, and forecasting: A systematic literature review. Agronomy 11, 1789. doi:. PMID: doi:10.3390/agronomy11091789
  5. Bonilla I., de Toda F. M., Martínez-Casasnovas J. A. (2015). Vine vigor, yield and grape quality assessment by airborne remote sensing over three years: Analysis of unexpected relationships in cv. Tempranillo. Span. J. Agric. Res. 13, e0903-e0903. doi:. PMID: doi:10.5424/sjar/2015132-7809
  6. Borgogno-Mondino E., Lessio A., Tarricone L., Novello V., de Palma L. A. (2018). Comparison between multispectral aerial and satellite imagery in precision viticulture. Precis. Agric. 19, 195 – 217. doi:. PMID: doi:10.1007/s11119-017-9510-0
  7. Campos J., García-Ruíz F., Gil E. (2021). Assessment of vineyard canopy characteristics from vigour maps obtained using UAV and satellite imagery. Sensors 21, 2363. doi:. PMID: doi:10.3390/s21072363
  8. Devaux N., Crestey T., Leroux C., Tisseyre B. (2019). Potential of Sentinel-2 satellite images to monitor vine fields grown at a territorial scale. OENO One 53, 51 – 58. doi: doi:10.20870/oeno-one.2019.53.1.2293
  9. Di Gennaro S. F., Dainelli R., Palliotti A., Toscano P., Matese A. (2019). Sentinel-2 validation for spatial variability assessment in overhead trellis system viticulture versus UAV and agronomic data. Remote Sens. 11, 2573. doi:. PMID: doi:10.3390/rs11212573
  10. Dobrowski S. Z., Ustin S. L., Wolpert J. A. (2002). Remote estimation of vine canopy density in vertically shoot‐positioned vineyards: determining optimal vegetation indices. Aust. J. Grape Wine Res. 8, 117 – 125. doi:. PMID: doi:10.1111/j.1755-0238.2002.tb00220.x
  11. Eichhorn K. W., Lorenz D. H. (1977). Phenological development stages of the grape vine. Nachrbl Dtsch. Pflanzenschutzd (Braunschweig) 29, 119 – 120.
  12. Esri Inc. (2025). ArcGIS Pro (Version 3.6). Redlands, CA: Environmental Systems Research Institute, Inc. https://pro.arcgis.com (Accessed April 21, 2025).
  13. European Space Agency (2015). Sentinel-2 user guide (Paris, France: ESA). Available online at: https://sentinels.copernicus.eu/web/sentinel/user-guides/sentinel-2-msi (Accessed September 20, 2025).
  14. Ferro M. V., Catania P., Micciche D., Pisciotta A., Vallone M., Orlando S. (2023). Assessment of vineyard vigour and yield spatio-temporal variability based on UAV high resolution multispectral images. Biosyst. Eng. 231, 36 – 56. doi:. PMID: doi:10.1016/j.biosystemseng.2023.06.001
  15. Filippetti I., Allegro G., Valentini G., Pastore C., Colucci E., Intrieri C. (2013). Influence of vigour on vine performance and berry composition of cv. Sangiovese (Vitis vinifera L.). OENO One 47, 21 – 33. doi: doi:10.20870/oeno-one.2013.47.1.1534
  16. Gatti M., Garavani A., Squeri C., Diti I., De Monte A., Scotti C.. (2022). Effects of intra-vineyard variability and soil heterogeneity on vine performance, dry matter and nutrient partitioning. Precis. Agric. 23, 150 – 177. doi:. PMID: doi:10.1007/s11119-021-09831-w
  17. Giovos R., Tassopoulos D., Kalivas D., Lougkos N., Priovolou A. (2021). Remote sensing vegetation indices in viticulture: A critical review. Agriculture 11, 457. doi:. PMID: doi:10.3390/agriculture11050457
  18. Guo W., Gu H., Adedeji O., Ghimire B. (2023). “ Advances in remote/aerial sensing of crop water status,” in Advances in sensor technology for sustainable crop production (Burleigh Dodds Science Publishing, Sawston, UK), 1 – 39. doi: doi:10.19103/as.2022.0107.01
  19. Hall A., Lamb D. W., Holzapfel B., Louis J. (2011). Within-season temporal variation in correlations between vineyard canopy and wine quality. Precis. Agric. 12, 103 – 117. doi:. PMID: doi:10.1007/s11119-010-9159-4
  20. Hu P., Sharifi A., Tahir M. N., Tariq A., Zhang L., Mumtaz F.. (2021). Evaluation of vegetation indices and phenological metrics using time-series MODIS data for monitoring vegetation change in Punjab, Pakistan. Water 13, 2550. doi:. PMID: doi:10.3390/w13182550
  21. Jiang Z., Huete A. R., Didan K., Miura T. (2008). Development of a two-band enhanced vegetation index without a blue band. Remote Sens. Environ. 112 (10), 3833 – 3845. doi: doi:10.1016/j.rse.2008.06.006
  22. Karn R., Hillin D., Helwi P., Scheiner J., Guo W. (2024). Assessing grapevine vigor as affected by soil physicochemical properties and topographic attributes for precision vineyard management. Sci. Hortic. 328, 112857. doi:. PMID: doi:10.1016/j.scienta.2024.112857
  23. Khaliq A., Comba L., Biglia A., Ricauda Aimonino D., Chiaberge M., Gay P. (2019). Comparison of satellite and UAV-based multispectral imagery for vineyard variability assessment. Remote Sens. 11, 436. doi:. PMID: doi:10.3390/rs11040436
  24. Kimmel T. M., Nielsen-Gammon J., Rose B., Mogil H. M. (2016). The weather and climate of Texas: A big state with big extremes. Weatherwise 69, 25 – 33. doi:. PMID: doi:10.1080/00431672.2016.1206446
  25. Lu L., Kuenzer C., Wang C., Guo H., Li Q. (2015). Evaluation of three MODIS-derived vegetation index time series for dryland vegetation dynamics monitoring. Remote Sens. 7, 7597 – 7614. doi:. PMID: doi:10.3390/rs70607597
  26. Matese A., Di Gennaro S. F. (2018). Practical applications of a multisensor UAV platform based on multispectral, thermal and RGB high-resolution images in precision viticulture. Agriculture 8, 116. doi:. PMID: doi:10.3390/agriculture8070116
  27. Matese A., Di Gennaro S. F., Orlandi G., Gatti M., Poni S. (2022). Assessing grapevine biophysical parameters from unmanned aerial vehicles hyperspectral imagery. Front. Plant Sci. 13, 898722. doi:. PMID: doi:10.3389/fpls.2022.898722
  28. Mucalo A., Matić D., Morić-Španić A., Čagalj M. (2024). Satellite solutions for precision viticulture: Enhancing sustainability and efficiency in vineyard management. Agronomy 14, 1862. doi:. PMID: doi:10.3390/agronomy14081862
  29. Ni W. (2024). Orbiting innovations: pioneering satellite technologies reshaping the future of vineyard management (Ithaca, NY, USA: Cornell University). Doctoral dissertation.
  30. Pastonchi L., Di Gennaro S. F., Toscano P., Matese A. (2020). Comparison between satellite and ground data with UAV-based information to analyse vineyard spatio-temporal variability. OENO One 54, 919 – 934. doi: doi:10.20870/oeno-one.2020.54.4.4028
  31. Python Software Foundation (2025). Python Official Website. Python.org. Available online at: https://www.python.org (Accessed August 12, 2025).
  32. R Core Team (2012). R: a language and environment for statistical computing (Vienna, Austria: R Foundation for Statistical Computing). Available online at: http://www.R-project.org/ (Accessed September 20, 2025).
  33. Reynolds A. G. (2022). “ Viticultural and vineyard management practices and their effects on grape and wine quality,” in Managing wine quality (Duxford, United Kingdom: Woodhead Publishing), 443 – 539.
  34. Rouse J., Haas R., Schell J., Deering D. (1973). Monitoring vegetation systems in the Great Plains with ERTS. Proceedings of the Third ERTS (Earth Resources Technology Satellite) Symposium. (Washington, DC, USA), 1, 309 – 317.
  35. Sassu A., Gambella F., Ghiani L., Mercenaro L., Caria M., Pazzona A. L. (2021). Advances in unmanned aerial system remote sensing for precision viticulture. Sensors 21, 956. doi:. PMID: doi:10.3390/s21030956
  36. Smart R. E. (1985). Principles of grapevine canopy microclimate manipulation with implications for yield and quality. A review. Am. J. Enol. Vitic. 36, 230 – 239. doi: doi:10.5344/ajev.1985.36.3.230
  37. Soil Survey Staff (2023). Web soil survey. Available online at: https://websoilsurvey.sc.egov.usda.gov/App/HomePage.htm (Accessed July 25, 2025).
  38. Sozzi M., Kayad A., Marinello F., Taylor J., Tisseyre B. (2020). Comparing vineyard imagery acquired from Sentinel-2 and unmanned aerial vehicle (UAV) platform. OENO One 54, 189 – 197. doi: doi:10.20870/oeno-one.2020.54.1.2557
  39. Sozzi M., Kayad A. G., Pezzuolo A., Sartori L., Dubbinic M., Marinello F. (2018). “ Comparison between different spatial resolution satellite images for viticulture ”, in: Proceedings of the European Conference on Agricultural Engineering, AgEng 2018 (Wageningen, The Netherlands: Wageningen University & Research).
  40. Steyn J., Aleixandre Tudo J., Aleixandre Benavent J. L. (2016). Grapevine vigour and within vineyard variability: A review. Int. J. Sci. Eng. Res. 7, 1056 – 1065. Available online at: https://riunet.upv.es/handle/10251/97767.
  41. Tardaguila J., Stoll M., Gutiérrez S., Proffitt T., Diago M. P. (2021). Smart applications and digital technologies in viticulture: A review. Smart Agric. Technol. 1, 100005. doi:. PMID: doi:10.1016/j.atech.2021.100005
  42. Tassopoulos D., Kalivas D., Giovos R., Lougkos N., Priovolou A. (2021). Sentinel-2 imagery monitoring vine growth related to topography in a protected designation of origin region. Agriculture 11, 785. doi:. PMID: doi:10.3390/agriculture11080785
  43. Taylor J. A., Bates T. R., Jakubowski R., Jones H. (2023). Machine-learning methods to identify key predictors of site-specific vineyard yield and vine size. Am. J. Enol. Vitic. 74, 0740013. doi: doi:10.5344/ajev.2022.22050
  44. Torrion J. A., Maas S. J., Guo W., Bordovsky J. P., Cranmer A. M. (2014). A three-dimensional index for characterizing crop water stress. Remote Sens. 6, 4025 – 4042. doi:. PMID: doi:10.3390/rs6054025
  45. Trought M. C., Bramley R. G. (2011). Vineyard variability in Marlborough, New Zealand: characterizing spatial and temporal changes in fruit composition and juice quality in the vineyard. Aust. J. Grape Wine Res. 17, 79 – 89. doi:. PMID: doi:10.1111/j.1755-0238.2010.00120.x
  46. Vance A. J. (2012). Impacts of crop level and vine vigor on vine balance and fruit composition in Oregon Pinot Noir (Corvallis, OR, USA: Oregon State University). Master Thesis.
  47. Wei H. E., Grafton M., Bretherton M., Irwin M., Sandoval E. (2023). Evaluation of the use of two-stage calibrated PlanetScope images and environmental variables for the development of the grapevine water status prediction model. Technol. Agron. 3, 6. doi: doi:10.48130/tia-2023-0006
  48. Winkler A. J. (1974). General viticulture (Berkley: University of California Press), 740.

Citation

Cudjoe, D. K., et al. (2026). Assessing spatial and temporal patterns of grapevine vigor and yield using satellite remote sensing. Frontiers in Horticulture, Volume 5. https://doi.org/
Received
Accepted
Published
Volume
5 - 2026

Copyright

Copyright © 2026 Cudjoe, Sun, Scheiner, Hillin, Helwi, Li and Guo.

License: https://creativecommons.org/licenses/by/4.0/

*Correspondence: Wenxuan Guo

Article metrics

Views
0
Downloads
0
Citations
0
Shares
0

Related articles

Original ResearchPerformance of smooth cayenne pineapple under tropical conditions: unravelling effects of postharvest duration and ripening stageKalolaini Koto Colaitiniyara, Shamal Shasang Kumar, Owais Ali Wani, Savenaca Cuquma, Mohammadi Begum Jeelani, Nouf Abdulrahman Alqahtani, Qasem Al-mdallalBrief Research ReportSentinel-2 NDVI as an indicator of apple orchard yieldN. Gasimzade, A. Radzhabov, I. SavinOriginal ResearchDevelopment of a robust, deployable loop-mediated isothermal amplification assay for rapid and specific detection of soft rot caused by PectobacteriumHassan Ali Tamur, Firas Ali Ahmed