Figure 1.
Mata research site overview map; red lines represent the field of view of the ground camera, with a black dot indicating the location of the camera; the green polygon represents the extent of the UAV mosaics; and the yellow polygon represents part of VENμS S01 tile (appearing in the inset map) which overlaps the research area.
Figure 1.
Mata research site overview map; red lines represent the field of view of the ground camera, with a black dot indicating the location of the camera; the green polygon represents the extent of the UAV mosaics; and the yellow polygon represents part of VENμS S01 tile (appearing in the inset map) which overlaps the research area.
Figure 2.
The spectral bands of the three sensors: UAV (16 bit), VENμS (16 bit), and ground photos (8 bit).
Figure 2.
The spectral bands of the three sensors: UAV (16 bit), VENμS (16 bit), and ground photos (8 bit).
Figure 3.
Data availability of ground photos, UAV and VENμS acquisitions in the research area, between December 2017 and January 2019.
Figure 3.
Data availability of ground photos, UAV and VENμS acquisitions in the research area, between December 2017 and January 2019.
Figure 4.
Identification of four different woody Mediterranean species at the individual plant level. Palestine oak in yellow, Terebinth in red, Mastic in green, and Carob in blue, over true color ground photo (a) and the 14 cm true color UAV mosaic (b), which were both acquired on 6 August 2018 at Mata. The two photos on the bottom right present an individual oak tree as acquired by the ground camera (c) and the UAV (d).
Figure 4.
Identification of four different woody Mediterranean species at the individual plant level. Palestine oak in yellow, Terebinth in red, Mastic in green, and Carob in blue, over true color ground photo (a) and the 14 cm true color UAV mosaic (b), which were both acquired on 6 August 2018 at Mata. The two photos on the bottom right present an individual oak tree as acquired by the ground camera (c) and the UAV (d).
Figure 5.
Data availability of VENμS for the TimeSat pixel-based analysis of seasonality. The first (NDVI) date was produced as an average of two neighbor clear dates (11 September 2018 and 29 October 2018).
Figure 5.
Data availability of VENμS for the TimeSat pixel-based analysis of seasonality. The first (NDVI) date was produced as an average of two neighbor clear dates (11 September 2018 and 29 October 2018).
Figure 6.
Mean values of the six vegetation indices for 12 species based on the ground photos after preprocessing. The
y-axis represents the standardized mean value of the vegetation index as a percent of the range of its values (12 species in 57 dates, i.e., one year, per index, min and max values in
Table S1), to ease comparison across the species and between the indices.
Figure 6.
Mean values of the six vegetation indices for 12 species based on the ground photos after preprocessing. The
y-axis represents the standardized mean value of the vegetation index as a percent of the range of its values (12 species in 57 dates, i.e., one year, per index, min and max values in
Table S1), to ease comparison across the species and between the indices.
Figure 7.
Standardized mean ExG values for the 12 species based on the ground photos after preprocessing (57 dates), and UAV data (11 dates, 14 cm resolution). The y-axis represents the index mean value as a percent of the range of the possible values for each sensor.
Figure 7.
Standardized mean ExG values for the 12 species based on the ground photos after preprocessing (57 dates), and UAV data (11 dates, 14 cm resolution). The y-axis represents the index mean value as a percent of the range of the possible values for each sensor.
Figure 8.
Standardized mean ExG values for the 12 species based on the UAV data (11 dates) after resampling to spatial resolutions of 14, 30, 125, and 500 cm. The y-axis represents the index mean value as a percent of the range of the possible values for each sensor.
Figure 8.
Standardized mean ExG values for the 12 species based on the UAV data (11 dates) after resampling to spatial resolutions of 14, 30, 125, and 500 cm. The y-axis represents the index mean value as a percent of the range of the possible values for each sensor.
Figure 9.
Scatter plots showing the correlations between the ExG mean values from the ground photos time series (x-axis) and from the UAV (11 dates), at pixel sizes (cm) of 14 (red), 30 (black), 125 (green), and 500 (blue). The colors of the linear trend lines correspond to the pixel sizes colors in the diagram. The asterisks represent the statistical significance of the correlations: *** p < 0.01, ** p < 0.01, * p < 0.05.
Figure 9.
Scatter plots showing the correlations between the ExG mean values from the ground photos time series (x-axis) and from the UAV (11 dates), at pixel sizes (cm) of 14 (red), 30 (black), 125 (green), and 500 (blue). The colors of the linear trend lines correspond to the pixel sizes colors in the diagram. The asterisks represent the statistical significance of the correlations: *** p < 0.01, ** p < 0.01, * p < 0.05.
Figure 10.
Standardized mean NDVI values of the 12 species based on the ground photos after preprocessing (57 dates), UAV data (11 dates, 14 and 500 cm), and VENμS acquisitions (47 dates, nominal resolution of 5.3 m at nadir, effective resolution of about 7.5 m due to the off-nadir acquisition angle of 29.7°). The y-axis represents the index mean value as a percent of the possible range of values.
Figure 10.
Standardized mean NDVI values of the 12 species based on the ground photos after preprocessing (57 dates), UAV data (11 dates, 14 and 500 cm), and VENμS acquisitions (47 dates, nominal resolution of 5.3 m at nadir, effective resolution of about 7.5 m due to the off-nadir acquisition angle of 29.7°). The y-axis represents the index mean value as a percent of the possible range of values.
Figure 11.
Standardized mean ExG (green), NDVI (black) and Relative Red (red) of Terebinth (n = 32), based on the ground photos after preprocessing. The left y-axis represents the index mean value as a percent of the range of values, and the daily minimum air temperature (C°, grey dots), recorded at Israeli Meteorological Service station #7220 (Netiv HaLamed-Heh, located about 7 km west of Mata). The right y-axis represents the daily precipitation (mm, blue bars), recorded at Israeli Meteorological Service station #246630 (Tzur Hadassah, located about 4.5 km east of Mata).
Figure 11.
Standardized mean ExG (green), NDVI (black) and Relative Red (red) of Terebinth (n = 32), based on the ground photos after preprocessing. The left y-axis represents the index mean value as a percent of the range of values, and the daily minimum air temperature (C°, grey dots), recorded at Israeli Meteorological Service station #7220 (Netiv HaLamed-Heh, located about 7 km west of Mata). The right y-axis represents the daily precipitation (mm, blue bars), recorded at Israeli Meteorological Service station #246630 (Tzur Hadassah, located about 4.5 km east of Mata).
Figure 12.
Correlations between ExG values of individual plants extracted from the ground photos and UAV images (11 dates) of the following species: Palestine Oak (A), Terebinth (B), Carob (C), Mastic tree (D), and mean ExG values of Herbaceous vegetation (all ROIs per date), detected by the same sensor (ground photos and UAV), at the individual plant level. Negative correlations range in the bluish background.
Figure 12.
Correlations between ExG values of individual plants extracted from the ground photos and UAV images (11 dates) of the following species: Palestine Oak (A), Terebinth (B), Carob (C), Mastic tree (D), and mean ExG values of Herbaceous vegetation (all ROIs per date), detected by the same sensor (ground photos and UAV), at the individual plant level. Negative correlations range in the bluish background.
Figure 13.
VENμS processed study area; (a) the location of Mata over the elevation base map. (b) bare surface, woody vegetation and herbaceous vegetation components. (c–e) a pixel-based analysis of phenological indices—start of season (SOS), maximum of season (MOS), and end of season (EOS), computed by TimeSat 3.3 software, for the time period between October 2018 and September 2019, using 62 cloud-free VENμS images. Grey areas represent urban, bare surface and agricultural land uses, which were masked out and not included in our analysis.
Figure 13.
VENμS processed study area; (a) the location of Mata over the elevation base map. (b) bare surface, woody vegetation and herbaceous vegetation components. (c–e) a pixel-based analysis of phenological indices—start of season (SOS), maximum of season (MOS), and end of season (EOS), computed by TimeSat 3.3 software, for the time period between October 2018 and September 2019, using 62 cloud-free VENμS images. Grey areas represent urban, bare surface and agricultural land uses, which were masked out and not included in our analysis.
Table 1.
Vegetation indices calculated for the ground camera (a–f), UAV and VENµS data (a–g).
Table 1.
Vegetation indices calculated for the ground camera (a–f), UAV and VENµS data (a–g).
| Vegetation Index | Formula | References |
---|
(a) | Green excess\Excess green; ExG |
2G − (R + B)
| [41] |
(b) | Normalized Difference Vegetation Index; NDVI | | [3] |
(c) | Green-Red Vegetation Index; GRVI | | [42] |
(d) | Relative red/RCC; Red chromatic coordinate | | |
(e) | Relative green/GCC; Green chromatic coordinate | | [43] |
(f) | Relative blue/BCC; Blue chromatic coordinate | | |
(g) | Normalized Difference Vegetation Index red-edge; NDVIre | | [44,45] |
Table 2.
List of the species analyzed with the number of individuals for each species per platform.
Table 2.
List of the species analyzed with the number of individuals for each species per platform.
Phenological Group | Species | Common Name | Number of ROIs |
---|
Ground Camera | UAV & VENμS |
---|
Evergreen | Pinus halepensis | Aleppo Pine | 11 | 42 |
Evergreen | Quercus calliprinos | Palestine Oak | 157 | 350 |
Evergreen | Pistacia lentiscus | Mastic tree, Lentisk | 74 | 324 |
Evergreen | Olea europaea | Olive tree | 18 | 49 |
Evergreen | Ceratonia siliqua | Carob, Locust tree | 21 | 64 |
Evergreen | Cistus salviifolius/creticus | Rock Rose | - | 94 |
Summer deciduous | Anagyris foetida | Mediterranean Stinkbush | 13 | 33 |
Summer deciduous | Calicotome villosa | Spiny Broom | 43 | 40 |
Summer semi-deciduous | Rhamnus lycioides | Palestine Buckthorn | 39 | 139 |
Summer semi-deciduous | Sarcopoterium spinosum | Prickly Burnet | 23 | 155 |
Winter deciduous | Pistacia palaestina | Terebinth | 32 | 47 |
Winter deciduous | Prunus dulcis | Common Almond | 89 | 120 |
| | Annual Herbaceous | 27 | 94 |
Total | | | 547 | 1551 |
Table 3.
UAV data acquisition dates and adjacent dates of ground photography.
Table 3.
UAV data acquisition dates and adjacent dates of ground photography.
Ground Photos | UAV |
---|
31 January 2018 | 1 February 2018 |
5 March 2018 | 4 March 2018 |
12 April 2018 | 8 April 2018 |
2 May 2018 | 3 May 2018 |
14 June 2018 | 13 June 2018 |
27 June 2018 | 27 June 2018 |
6 August 2018 | 6 August 2018 |
24 September 2018 | 25 September 2018 |
20 November 2018 | 22 November 2018 |
25 December 2018 | 24 December 2018 |
22 January 2019 | 21 January 2019 |
Table 4.
Pearson correlation coefficients of six Vegetation Indices mean values for each species per date (11 dates), between values derived from the ground camera data (after pre-processing), and those from the UAV data (spatial resolution of 14 cm). The asterisks indicate the statistical significance as follows: * p < 0.05, ** p < 0.01, *** p < 0.001. H0: R2 = 0.
Table 4.
Pearson correlation coefficients of six Vegetation Indices mean values for each species per date (11 dates), between values derived from the ground camera data (after pre-processing), and those from the UAV data (spatial resolution of 14 cm). The asterisks indicate the statistical significance as follows: * p < 0.05, ** p < 0.01, *** p < 0.001. H0: R2 = 0.
Species | ExG | GRVI | NDVI | Relative Blue | Relative Green | Relative Red |
---|
Aleppo Pine | 0.65 * | −0.19 | 0.31 | 0.70 * | 0.51 | −0.51 |
Terebinth | 0.95 *** | 0.49 | 0.47 | 0.54 | 0.58 | 0.24 |
Palestine Oak | 0.88 *** | −0.08 | 0.83 ** | 0.63 * | 0.16 | 0.06 |
Mastic | 0.87 *** | −0.13 | 0.72 ** | 0.40 | −0.14 | 0.19 |
Palestine Buckthorn | 0.91 *** | 0.55 | 0.23 | 0.82 ** | 0.61 * | 0.82 ** |
Olive | 0.72 * | 0.31 | 0.27 | 0.39 | −0.13 | 0.75 ** |
Carob | 0.76 | −0.15 | 0.47 | 0.64 * | 0.33 | −0.36 |
Prickly Burnet | 0.91 *** | 0.96 *** | 0.86 *** | 0.62 * | 0.80 ** | 0.81 |
Annual Herbaceous | 0.93 *** | 0.95 *** | 0.96 *** | 0.71 * | 0.88 *** | 0.86 *** |
Mediterranean Stinkbush | 0.87 *** | 0.70 ** | 0.80 ** | 0.74 ** | 0.70 * | 0.50 |
Spiny Broom | 0.91 *** | 0.85 *** | 0.79 ** | 0.61 * | 0.63 * | 0.68 * |
Almond | 0.92 *** | 0.90 *** | 0.77 ** | 0.92 *** | 0.91 *** | 0.38 |
Table 5.
Pearson correlation coefficients of seven spectral indices mean value for each species per date (10 dates), between values derived from the UAV data and those from VENμS. The asterisks indicate the statistical significance as follows: * p < 0.05, ** p < 0.01, *** p < 0.001. H0: R2 = 0.
Table 5.
Pearson correlation coefficients of seven spectral indices mean value for each species per date (10 dates), between values derived from the UAV data and those from VENμS. The asterisks indicate the statistical significance as follows: * p < 0.05, ** p < 0.01, *** p < 0.001. H0: R2 = 0.
Spatial Resolution (cm) | Aleppo Pine | Terebinth | Palestine Oak | Mastic Tree | Palestine Buckthorn | Olive Tree | Carob | Annual Herbaceous | Mediterranean Stinkbush | Spiny Broom | Common Almond | Prickly Burnet |
---|
ExG | 0.02 | −0.75 * | −0.21 | −0.43 | 0.09 | −0.08 | −0.36 | 0.97 *** | 0.74 * | 0.66 * | −0.29 | 0.90 *** |
GRVI | 0.67 * | −0.42 | 0.60 | 0.59 | 0.68 * | 0.67 * | 0.58 | 0.93 *** | 0.92 *** | 0.86 ** | −0.08 | 0.95 *** |
NDVI | 0.74 * | −0.32 | 0.60 | 0.67 * | 0.69 * | 0.65 * | 0.60 | 0.92 *** | 0.85 ** | 0.87 ** | 0.19 | 0.90 *** |
NDVIre | 0.90 *** | −0.53 | 0.51 | 0.70 * | 0.76 ** | 0.70 * | 0.69 * | 0.93 *** | 0.83 ** | 0.88 *** | 0.32 | 0.88 *** |
R Blue | −0.38 | −0.10 | −0.27 | 0.27 | 0.39 | 0.03 | −0.54 | −0.65 * | −0.14 | 0.14 | 0.61 | 0.45 |
R Green | 0.52 | −0.27 | 0.63 * | 0.59 | 0.71 * | 0.68 * | 0.52 | 0.91 *** | 0.91 *** | 0.88 *** | 0.02 | 0.95 *** |
R Red | 0.69 * | −0.50 | 0.57 | 0.60 | 0.73 * | 0.64 * | 0.58 | 0.91 *** | 0.94 *** | 0.84 ** | 0.03 | 0.95 *** |
Table 6.
Summary of the strongest Pearson correlations coefficients of the spectral indices between values derived from the ground camera data (after pre-processing), and those from the UAV data, after resampling to different pixel sizes. For each cell in the table, the correlation coefficient is shown only for the vegetation index with the strongest correlation: ExG in black, GRVI in orange, NDVI in purple, Relative Blue in blue, Relative Green in green, and Relative Red in red. The asterisks indicate the statistical significance as follows: * p < 0.05, ** p < 0.01, *** p < 0.001. H0: R2 = 0.
Table 6.
Summary of the strongest Pearson correlations coefficients of the spectral indices between values derived from the ground camera data (after pre-processing), and those from the UAV data, after resampling to different pixel sizes. For each cell in the table, the correlation coefficient is shown only for the vegetation index with the strongest correlation: ExG in black, GRVI in orange, NDVI in purple, Relative Blue in blue, Relative Green in green, and Relative Red in red. The asterisks indicate the statistical significance as follows: * p < 0.05, ** p < 0.01, *** p < 0.001. H0: R2 = 0.
Spatial Resolution (cm) | Aleppo Pine | Terebinth | Palestine Oak | Mastic Tree | Palestine Buckthorn | Olive Tree | Carob | Annual Herbaceous | Mediterranean Stinkbush | Spiny Broom | Common Almond | Prickly Burnet |
---|
14 | 0.70 * | 0.95 *** | 0.88 *** | 0.87 *** | 0.91 *** | 0.75 ** | 0.76 | 0.96 *** | 0.87 *** | 0.91 *** | 0.92 *** | 0.96 *** |
30 | 0.70 * | 0.95 *** | 0.87 *** | 0.86 *** | 0.90 *** | 0.75 ** | 0.75 | 0.96 *** | 0.87 *** | 0.91 *** | 0.92 *** | 0.96 *** |
125 | 0.66 * | 0.93 *** | 0.87 *** | 0.87 *** | 0.85 *** | 0.73 ** | 0.72 * | 0.96 *** | 0.86 *** | 0.90 *** | 0.89 *** | 0.96 *** |
500 | 0.52 | −0.71 | 0.81 ** | 0.83 ** | 0.73 ** | 0.68 * | −0.58 | 0.95 *** | 0.71 * | 0.89 *** | 0.70 * | 0.94 *** |
Table 7.
An overview of the classification accuracy assessment results (average producer’s accuracies) of the ExG time series, based on the UAV images, after resampling them to 4 different spatial resolutions.
Table 7.
An overview of the classification accuracy assessment results (average producer’s accuracies) of the ExG time series, based on the UAV images, after resampling them to 4 different spatial resolutions.
Species | Pixel Size (cm) |
---|
14 | 30 | 125 | 500 |
---|
Aleppo Pine | 80% | 77% | 71% | 11% |
Terebinth | 64% | 50% | 44% | 12% |
Palestine Oak | 63% | 73% | 61% | 21% |
Mastic | 72% | 72% | 50% | 23% |
Palestine Buckthorn | 48% | 53% | 34% | 9% |
Olive | 80% | 78% | 55% | 27% |
Carob | 48% | 42% | 36% | 35% |
Rock Rose | 71% | 73% | 66% | 54% |
Prickly Burnet | 92% | 90% | 78% | 52% |
Annual Herbaceous | 98% | 97% | 82% | 49% |
Mediterranean Stinkbush | 76% | 66% | 58% | 23% |
Spiny Broom | 56% | 47% | 39% | 19% |
Almond | 87% | 80% | 67% | 51% |
total accuracy | 0.72 | 0.69 | 0.57 | 0.23 |
Kappa | 0.70 | 0.67 | 0.53 | 0.19 |
Table 8.
Matrix of Spearman’s rank correlation coefficients of the phenological indices (SOS, MOS, EOS) and the environmental factors after normalization.
n = 1,716,915 pixels per variable, including only areas of Mediterranean vegetation, as shown in
Figure 13. The phenological indices were calculated for the time period between October 2018 and September 2019, using 62 cloud-free VENμS images.
Table 8.
Matrix of Spearman’s rank correlation coefficients of the phenological indices (SOS, MOS, EOS) and the environmental factors after normalization.
n = 1,716,915 pixels per variable, including only areas of Mediterranean vegetation, as shown in
Figure 13. The phenological indices were calculated for the time period between October 2018 and September 2019, using 62 cloud-free VENμS images.
| SOS | MOS | EOS | Elevation | Aspect | Slope | % Woody |
---|
MOS | −0.140 | - | | | | | |
EOS | −0.175 | 0.863 | - | | | | |
Elevation | −0.290 | 0.272 | 0.326 | - | | | |
Aspect | −0.054 | −0.363 | −0.282 | 0.133 | - | | |
Slope | −0.058 | 0.362 | 0.349 | 0.332 | −0.073 | - | |
% woody | −0.248 | 0.322 | 0.318 | −0.093 | −0.223 | 0.019 | - |
% herbaceous | 0.331 | −0.500 | −0.515 | −0.207 | 0.032 | −0.222 | −0.644 |