ISSN 2953-6367  
http://revistainvestigo.com  
Julio - diciembre 2026  
Vol. 7 No. 20, PP. 869-889  
https://doi.org/10.56519/k6y27q77  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL  
ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL  
CON UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
SPATIOTEMPORAL ASSESSMENT OF GREEN MANURE  
GROWTH THROUGH MULTISPECTRAL UAV REMOTE SENSING,  
GIS AND SOIL PHYSICOCHEMICAL ANALYSIS  
Hernán Eriberto Chamorro Sevilla1, Rodrigo Ernesto Salazar López2, Nanci Margarita Inca  
Chunata3, Keylly Marcela Chávez Inca4  
{hernan.chamorro@espoch.edu.ec1 , rodrigo.salazar@espoch.edu.ec2 , ninca@espoch.edu.ec3, keylly.chavez@spoch.edu.ec4}  
Fecha de recepción: 15/06/2026  
/ Fecha de aceptación: 06/07/2026  
/ Fecha de publicación: 08/07/2026  
ABSTRACT: This study evaluated the spatiotemporal dynamics of a green manure mixture  
(Vicia sativa L. and Avena sativa L.) under high-Andean conditions using multispectral UAV  
remote sensing, GIS, and soil physicochemical analysis. Field experiments were conducted at  
an altitude of 2,850 m above sea level using DJI Mavic 3 Multispectral UAV. Data were collected  
in March and May 2026. Correlation analysis was performed between vegetation indices (NDVI,  
SAVI, and NDRE) and soil physicochemical properties, including pH, organic carbon,  
phosphorus, potassium, and soil texture. Laboratory analyses showed that the soil was slightly  
alkaline, with a sandy loam texture but low concentrations of organic carbon and available  
phosphorus. Multispectral imagery revealed a considerable increase in canopy vigor and  
biomass between the two monitoring dates. SAVI effectively minimized soil background effects  
during the early growth stage, whereas NDVI showed greater temporal variability associated  
with biomass development. In contrast, NDRE remained relatively stable, indicating a more  
uniform chlorophyll distribution and physiological condition of the crop. The integration of  
multispectral UAV imagery, GIS, and soil physicochemical analysis proved to be an effective  
approach for assessing the spatiotemporal development of green manure crops under high-  
Andean conditions. This integrated methodology enabled the identification of spatial variability  
in crop development associated with soil physicochemical properties, providing valuable  
information to support site-specific crop management and sustainable agricultural practices.  
1Escuela Superior Politécnica de Chimborazo, Ecuador, https://orcid.org/0000-0002-8531-7116; +5930996487535  
2Escuela Superior Politécnica de Chimborazo (ESPOCH) Sede Orellana, Ecuador, https://orcid.org/0000-0001-6194-1638;  
+593999444777  
3Escuela Superior Politécnica de Chimborazo (ESPOCH, Ecuador, https://orcid.org/0000-0001-6703-0878; +5930984672267  
4Escuela Superior Politécnica de Chimborazo (ESPOCH) Sede Orellana, Ecuador, https://orcid.org/0000-0003-0780-6704;  
+593998603125  
869  
Revista Científica Multidisciplinaria InvestiGo  
Riobamba – Ecuador  
Cel: +593 97 911 9620  
revisinvestigo@gmail.com  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
Keywords: Precision agriculture, green manure, UAV remote sensing, vegetation indices, soil  
physicochemical analysis, Andean agriculture  
INTRODUCCIÓN  
Increased usage of sustainable agriculture has resulted in an increasing trend in practices that  
increase soil quality through reducing the reliance on chemical fertilizers and other external  
inputs. The use of green manures is among the most efficient means of replenishing soil fertility,  
nutrient provision, improving soil structure, and promoting sustainability. Incorporation of green  
manures into the process of agricultural production is known to provide for organic matter  
addition, improved soil aggregate formation, increased water infiltration and storage, and  
reduced soil erosion (1-4). Apart from improving the physical and chemical characteristics of soils,  
the use of green manures stimulates biological activity through creating conditions for microbes  
involved in nutrients mineralization and organic matter decomposition (1-4).  
Leguminous green manures are of special importance due to the ability to develop symbiosis with  
nitrogen-fixing bacteria, allowing for incorporation of atmospheric nitrogen into the soil-plant  
system. Combining leguminous green manures with grass like oats (Avena sativa L.) results in  
greater efficiency of nutrient recycling, better protection of the soil surface, and maintenance of  
the proper C: N ratio during decomposition of the green manure materials (5-8).  
However, the performance of the green manure crops shows considerable spatial variability  
across agricultural plots. The variability of soil parameters, such as its texture, organic matter  
content, nutrient status, moisture and others causes the uneven growth and biomass formation  
of the crops in distinct parts of agricultural fields. Traditional field evaluations are mostly based  
on biomass harvesting and observation; although such methods allow us to obtain valuable  
information about the state of crops, they are only certain parts of field conditions. These  
methods are labor and time-intensive, and they lack the spatial variability of crops at the field  
level especially during early growth stage (9–11).  
Precision agriculture technologies including the remote sensing allow to significantly increase the  
accuracy of evaluation of crop condition by monitoring from above. Among remote sensing  
techniques, the use of UAVs with multispectral imaging sensors is becoming increasingly popular  
because of the important level of flexibility and resolution of the imagery provided by such  
equipment. In contrast to the satellite platforms, the use of UAVs allows the repetition of the  
observations under the favorable local weather conditions and provide the dataset that can be  
used for field-scale analysis and are appropriate for the short cycle crops and cover crops (12–  
15).  
Multispectral imagery makes it possible to derive vegetation indices characterizing physiological  
status of crops because of spectral reflectance data. The Normalized Difference Vegetation Index  
(NDVI) is one of the most widely used spectral indices for estimating canopy vigor, biomass  
accumulation, and photosynthetic activity in agricultural systems and has shown reliable  
performance under heterogeneous environmental conditions, including high-Andean  
870  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
agroecosystems (3,22,25). The Normalized Difference Red Edge Index (NDRE) is more sensitive to  
variations in chlorophyll content and plant nutritional status, particularly after canopy closure,  
making it suitable for monitoring crop physiological condition in precision agriculture (23).  
Likewise, the Soil-Adjusted Vegetation Index (SAVI) minimizes the influence of exposed soil during  
the early stages of crop establishment, improving vegetation discrimination when vegetation  
cover is still incomplete (2,23). Previous studies conducted in Andean environments have  
demonstrated that the combined interpretation of NDVI, NDRE, and SAVI provides a more  
comprehensive assessment of crop development than the use of a single vegetation index by  
integrating information related to biomass production, physiological status, and soil background  
effects (22–25).  
The value of vegetation indices increases considerably by incorporating these indicators into GIS  
systems. The ability to perform spatial analysis makes it possible to integrate vegetation maps  
with soil properties obtained in laboratory experiments, digital terrain models, and other  
environmental factors to reveal relationships between soil characteristics and spatial pattern of  
crop growth. This provides for creation of management zones, support of variable rate application  
in agriculture, and precision agriculture. Besides, the integration of multispectral imagery and  
physicochemical analysis of soils allows revealing how soil heterogeneity affects vegetation and  
crop development (21–24).  
Whereas UAV-assisted remote sensing techniques have widely been investigated in commercial  
crops, fewer studies have looked at green manures grown under high altitude Andean  
environmental conditions. Fast growing species like common vetch (Vicia sativa L.) and oat (Avena  
sativa L.), which respond highly to changes in soil properties, serve well as model crops to assess  
the vegetation indices-soil properties interaction. The investigation into this interaction is  
necessary to enhance fertilizer application strategies, biomass production and sustainable  
agriculture.  
This research project was intended to use UAV-assisted multispectral remote sensing data along  
with GIS and soil physicochemical property analysis to assess the spatiotemporal development of  
a green manure crop developed under Andean environmental conditions. The second aspect of  
the research was to assess the interaction between vegetation indices and soil properties.  
MATERIALES Y MÉTODOS  
2.1 Study area  
871  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
Figure 1. Location of the study area  
The study was conducted at the experimental field of the Precision Agriculture Club, Faculty of  
Natural Resources, Escuela Superior Politécnica de Chimborazo (ESPOCH), which is situated at  
Riobamba, Province of Chimborazo, Ecuador. The experimental field covers an area of around  
9,138.79 m² and is found at the Universal Transverse Mercator (UTM) coordinates 758134.34 E,  
9817336.70 N (WGS 84, Zone 17S). Its average altitude is 2,850 m above sea level. The  
environment has an equatorial high-mountain, semi-humid mesothermal climate and has an  
average annual temperature of about 13.6 °C. It has high solar radiation and low nocturnal  
temperatures. All these characteristics are typical of the highland Andes of central Ecuador and  
make an ideal setting for evaluating the establishment and early growth of green manures in  
mountain agriculture.  
The experimental planting consisted of a mixed seeding of common vetch (Vicia sativa L.) and oat  
(Avena sativa L.), shown on 6 February 2026. These crops were chosen because legumes fix  
nitrogen biologically and cereals produce rapid biomass growth and improve soil cover.  
2.2 Soil Sampling and Laboratory Analyses  
Soil sampling was performed prior to crop establishment as a means of assessing the first  
physicochemical conditions of the experimental field. Composite soil samples were taken using  
872  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
an established soil sampling protocol and were analyzed in the Soil Laboratory of the Faculty of  
Natural Resources, ESPOCH, using standard methods of soil analysis.  
The variables analyzed included soil pH, electrical conductivity (EC), organic carbon, soil moisture  
content, available phosphorous (P), exchangeable potassium (K), soil bulk density, soil particle  
density, soil texture, and soil structure. The selection of the above-mentioned variables was based  
on the criteria that they are the most indicative of soil fertility and nutrient content, water  
retention ability, and physical characteristics affecting crop establishment.  
Laboratory data was later integrated into a geospatial database for comparative analysis with soil  
variables and vegetation indexes obtained via multispectral UAV imagery.  
2.3 UAV Data Acquisition and Image Processing  
The monitoring of crops was done through a DJI 3 Multispectral (M3M RTK) UAV fitted with a  
multispectral camera and RTK positioning which ensures centimeter level accuracy in positioning  
during data acquisition. UAV flights were done twice during the establishment of the green  
manure crop on 21 March and 26 May 2026 allowing the evaluation of changes of development  
of the crop during time.  
Vegetation indices were calculated from the multispectral bands acquired by the DJI Mavic 3  
Multispectral sensor after radiometric calibration and orthomosaic generation. NDVI, SAVI, and  
NDRE were selected because they provide complementary information on canopy vigor, soil  
background effects, and crop physiological status during crop establishment. Table 1 summarizes  
the analytical equations and spectral bands used to calculate each vegetation index.  
Table 1. Vegetation indices, analytical equations, and spectral bands used in this study  
Vegetation  
index  
Analytical equation  
Spectral Bands (DJI Mavic 3  
Multispectral)  
Main application  
푁퐼푅 − 푅푒푑  
(
)
푁퐼푅 840 ± 26푛푚  
Estimation of canopy vigor,  
biomass accumulation, and  
photosynthetic activity  
NDVI  
(
)
(
/푅푒푑 650  
± 16푛푚  
푁퐼푅 + 푅푒푑  
)
(
)
(
)
푁퐼푅 840 ± 26푛푚 /푅푒푑 650 ± 16푛푚  
SAVI  
Reduction of soil background  
influence during early crop  
development  
(
)
푁퐼푅 − 푅푒푑  
(
)
(
1 + 퐿 ),  
(
)
푁퐼푅 + 푅푒푑 + 퐿  
푤ℎ푒푟푒 퐿 = 0.5  
푁퐼푅 − 푅푒푑퐸푑푔푒  
푁퐼푅 + 푅푒푑퐸푑푔푒  
(
)
(
푁퐼푅 840 ± 26푛푚 /푅푒푑퐸푑푔푒 730  
NDRE  
Assessment of chlorophyll  
content and crop physiological  
condition  
(
)
)
± 16푛푚  
The flight plan was programmed with the use of autonomous UAV flying software, and the flights  
were flown at the height of 40 m above ground level, 80% forward overlap, 70% side overlap and  
average ground speed of 6 m s⁻¹. This ensured sufficient image overlaps to allow accurate  
873  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
photogrammetric reconstruction while maintaining high spatial resolution appropriate for  
precision agriculture purposes.  
Image processing was performed using Pix4Dmapper version 4.5.6. The image processing  
workflow implemented in Pix4Dmapper consisted of: (i) radiometric calibration using the DJI  
Mavic 3 Multispectral Reflectance Calibration Panel (CRP); (ii) initial image processing and  
automatic aerial triangulation; (iii) camera optimization; (iv) dense point cloud generation; (v)  
digital surface model (DSM) construction; (vi) multispectral orthomosaic generation for each  
spectral band; and (vii) export of calibrated reflectance orthomosaics as GeoTIFF files for  
subsequent analysis in ArcGIS Pro. During image acquisition, the integrated Sunshine Sensor was  
enabled to record real-time incident solar irradiance, allowing automatic correction of  
illumination variability between flight campaigns. Radiometric calibration converted the raw  
digital numbers into surface reflectance values, improving the radiometric consistency and  
comparability of the multispectral orthomosaics. Vegetation indices (NDVI, SAVI, and NDRE) were  
subsequently calculated in ArcGIS Pro using the Raster Calculator, and the resulting maps were  
integrated with the soil physicochemical data for spatial analysis.  
2.4 Vegetation Indices  
Three vegetation indices were used to find the development of crops throughout the entire  
monitoring period. NDVI index was used to measure the vigor of the plants, green biomass, and  
photosynthesis. The calculation of NDRE index allowed estimating the level of chlorophyll  
concentration and nutrient condition of the crop, especially when in advanced vegetative stage,  
since NDVI index can become saturated. SAVI index was considered for decreasing the influence  
of soil on the results of measurement and providing more exact estimation of plant performance  
with partial canopy coverage. The simultaneous use of these indices made it possible to conduct  
analysis of the physiological state of the crop, including measurements of biomass, chlorophyll  
content, and soil-adjusted vegetation index.  
2.5 Spatial and Statistical Analysis  
Spatial analyses were conducted using ArcGIS Pro software through the combination of  
vegetation indices maps and laboratory results of soil physicochemical properties. Thus, spatial  
variability of plant development was visualized and possible correlations between vegetation  
indices and soil characteristics could be found within the experimental field.  
Descriptive statistical analysis was done using R. Mean value, minimum value, maximum value,  
standard deviation, quartile values, and coefficient of variation were calculated for each  
vegetation index and soil parameter. Boxplots and histograms were created to evaluate changes  
in distribution of vegetation indices over time between the two periods of observations. The  
combination of descriptive statistical analysis, spatial analysis, and laboratory data allows  
evaluating the effect of soil physicochemical properties on early development of green manure  
crops under precision agriculture conditions.  
874  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
RESULTADOS  
3.1 Soil Physic-Chemical Characterization  
The physic-chemical characterization of the soil used in this experiment created an environmental  
benchmark which was needed to understand the multi-spectral responses of the green manure  
mixture during the course of the experiment. Analysis in the laboratory showed that the area  
under study had favorable environmental conditions for crop growth; however, variations in soil  
nutrient distribution may cause certain effects on the first vegetative growth. Combining  
laboratory analysis of the soil properties with UAV based multi-spectral data can help interpret  
the resulting spectra in relation to agriculture.  
Table 2. Physicochemical properties of the experimental soil before green manure establishment.  
So, soil analysis provided information about slightly alkaline reaction of the soil together with very  
low levels of soil electrical conductivity, which proves that there were no such factors as soil  
acidity or salinity limiting the possibility of crop establishment. Since pH of the soil belongs to the  
range where legumes and cereals can develop, it allows concluding about the good nutrient status  
of the soil and microorganisms taking part in the decomposition of the organic matter. Likewise,  
low electrical conductivity values confirm that there are no salts limiting crop establishment via  
causing osmotic stress during seed germination or root growth. Therefore, spectral heterogeneity  
seen in the UAV images cannot be related to the effects of salinity but should be caused by natural  
differences in the state of soil fertility and crops.  
As for the fertility parameters, organic carbon and available phosphorus are the major nutrient  
limitations of the experimental field. Organic carbon is important in terms of supporting soil  
structure, soil moisture, cation exchange capacity, and microorganisms, so its low level shows low  
availability of mineralizable organic matter able to provide nutrients at the stage of crop  
establishment. On the other hand, phosphorus is a principal element involved in root growth and  
energy transfer, so its low availability can explain the differences in crops' development seen  
while monitoring. Despite there being no nutrition-related problems detected in the  
875  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
experimental field, local phosphorus deficiency is probably responsible for the heterogeneous  
vegetation indices obtained after the first UAV flight.  
Potassium in an exchangeable form occurred at higher levels in the research area. Potassium  
regulates stomata opening, activation of enzymes, transport of carbohydrates, and water-use  
efficiency, all of which have direct relations to vegetative development and physiology of the  
plant in general. Potassium presence might be one of the factors that helped crops grow despite  
moderate nutrient deficiency in organic carbon and phosphorus. It is possible that this nutrient  
balance led to vegetation growing stronger over second observations.  
Besides, physical attributes of the soil were favorable for crop growth. Soil bulk density pointed  
to low compaction of soil and therefore good pore continuity for root system development and  
gas exchange. The same concerns particle density, which is consistent with mineral agricultural  
soils without any structural deterioration. These properties allow for unrestricted roots  
development and soil exploration as well as sufficient oxygen diffusion.  
According to the results of the texture analysis, sandy loam and sandy clay loam soil classes  
prevailed in the experimental field. Such textural classes provide a good balance between  
drainage, aeration, and water retention, which means that the risks of waterlogging can be  
avoided but at the same time soil keeps enough moisture needed for successful crop  
establishment. Even though coarse soils have less capability to retain nutrients than clays, they  
allow for fast drainage and reduce runoff, providing favorable physical conditions for crop  
emergence. Therefore, the results of the textural analysis may explain the uniformity of the  
canopy coverage recorded during the second UAV survey.  
Taking into account the interpretation of both chemical and physical properties of soils, it can be  
stated that crop establishment took place in such an environment that had no physical  
restrictions; thus, nutrient variability became the main condition that could produce spatial  
heterogeneity of vegetation development. This conclusion is rather important since it gives an  
agricultural explanation of the variability obtained through multispectral images. Low vegetation  
index values areas, therefore, may be associated with the nutrient availability in the area and not  
with poor physical soil conditions. It is clear that the results of laboratory analyses allow  
integrating remote sensing and laboratory measurements and making the interpretation of the  
spatial vegetation variability possible.  
From the point of view of precision agriculture, the preliminary soil analyses provide the basic  
framework for the interpretation of further crop vigor, chlorophyll and canopy development  
changes. Thus, the laboratory results are not the independent part of the study but the linking  
element between soil fertility and multispectral imagery that allows interpreting vegetation  
indices as crop responses to the existing soil conditions.  
The lab characterization forms the foundation for interpreting the evolution of spectra recorded  
over the monitoring period. While the physicochemical characterization describes the  
environmental conditions that prevailed during the growth of the crop, the vegetation indices  
determined by the UAV record the effects of those conditions on the crop at spatial and temporal  
876  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
levels. The following section evaluates the effects through the analysis of NDVI and SAVI, two  
vegetation indices that capture the effects of the environmental conditions during the  
establishment stage of the crop.  
The physicochemical characterization revealed that the environmental conditions in the  
experimental field were suitable for growing the crop despite variations in nutrient availability.  
To understand the influence of these environmental conditions on the growth of the crop, an  
analysis of the temporal evolution of vegetation vigor was conducted using NDVI and SAVI.  
3.2 Spatiotemporal dynamics of vegetation vigor (NDVI and SAVI)  
Temporal analysis of vegetation vigor shows a significant improvement in crop growth due to the  
gradual growth of the green manure mixture at the beginning of the growing season. Both NDVI  
and SAVI show an increasing trend in the canopy vigor in March-May. At the same time, each of  
the indices considers a different aspect of crop growth, since NDVI reflects mainly the increment  
of photosynthetically active biomass, while SAVI gives a better insight into the growth of the crops  
during the period of their establishment due to the elimination of the impact of soil brightness  
on results.  
The descriptive statistics provided in descriptive statistics prove the clear tendency of increase in  
the central tendency of the vegetation indices between the monitoring dates. The increase in the  
mean and median of vegetation indices proves the general growth of vegetation vigor for most  
of the pixels as the crop grows. This assumption can be additionally supported by the fact that  
the upper quartile increases as well, which means that the crops become more vigorous not only  
in the most productive regions but also in the intermediate ones.  
Table 3. Descriptive statistics of NDVI and SAVI obtained from UAV imagery during March and May.  
Though sharing a common trajectory with respect to time, there were slight differences in the  
way the two indicators responded to vegetation development, as reflected in their respective  
statistical distributions. The NDVI indicator experienced a continual increase in both its mean and  
median values over time, implying continued increase for vegetation cover while showing the  
crops. This was accompanied by a decreasing coefficient of variation, signifying the uniformity of  
biomass production in the area as canopy closure progressed. This decrease in the relative  
variability implies that the portions where crop emergence lagged in March managed to develop  
their vigor to levels like those of the already developed areas by May.  
877  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
3.3 Relationship between soil physicochemical properties and vegetation indices  
To complement the descriptive analysis, Spearman's rank correlation coefficient was applied to  
evaluate the relationships between soil physicochemical properties and vegetation indices. This  
non-parametric approach was selected because of the limited sample size (n = 6), making it more  
appropriate than parametric correlation methods. This non-parametric approach was selected  
because of the limited sample size (n = 6), making it more appropriate than parametric correlation  
methods. A strong negative correlation was found between soil organic carbon and both NDVI  
and SAVI during the first monitoring campaign (ρ = −0.926, p = 0.008), indicating that areas with  
higher organic carbon content exhibited lower spectral responses during the early establishment  
stage of the green manure crop. Likewise, electrical conductivity showed a strong negative  
correlation with NDRE in March (ρ = −0.829, p = 0.042). No statistically significant relationships (p  
> 0.05) were detected between the remaining soil properties and the vegetation indices  
evaluated. These findings suggest that soil properties may have influenced spectral responses  
during the initial crop establishment stage. This relationship became less evident after canopy  
closure, when vegetation became more spatially homogeneous.  
Table 4. Spearman correlation coefficients between soil physicochemical properties and vegetation indices.  
Soil property  
pH  
Conductivity  
Organic carbon  
Moisture  
NDVI (March)  
0.486  
NDVI (May)  
0.200  
SAVI (March)  
0.486  
SAVI (May)  
0.200  
NDRE (March)  
0.029  
NDRE (May)  
-0.257  
0.600  
-0.371  
-0.926*  
-0.600  
0.371  
0.123  
0.029  
-0.371  
-0.926*  
-0.600  
0.371  
0.123  
0.029  
-0.829*  
-0.617  
-0.029  
0.463  
0.086  
Phosphorus  
0.714  
0.543  
0.714  
0.543  
0.600  
0.486  
Potassium  
-0.257  
0.143  
-0.257  
0.143  
-0.714  
0.486  
It would be easier to visualize the dynamics through box plots compared to descriptive statistics  
alone. For the first monitoring period, the interquartile range covered a wide array of vegetation  
states, signifying heterogeneity of the crop emergence and the presence of seedlings alongside  
sizable soil exposure. The relatively low value of the media, together with many data points below  
the first quartile, signify an early establishment phase where vegetation had not yet developed  
enough to cover the whole soil surface.  
Figure 2. NDVI boxplots correspond to the March and May UAV surveys.  
878  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
It should be noted that during the second monitoring session, the median value in the NDVI  
boxplot was increased, and the upper quartile also showed positive dynamics. This is explained  
by the continuous biomass growth at the monitoring site and the gradual increase in homogeneity  
of canopy cover of the experimental area. Even though there were still some small values on the  
plot, their amount significantly decreased which shows that most parts of the plot were able to  
develop vegetation. Decrease in the relative dispersion of NDVI values at the second monitoring  
session is also indicative of crop development becoming more even over time.  
In addition, the NDVI histograms also confirm this conclusion. Thus, the histogram for March is  
characterized by high dispersion of low-positive values as well as a considerable number of pixels  
related to sparse vegetation and exposed soils. However, in May, the histogram showed the trend  
towards higher values with a significant decrease in the proportion of zero pixels.  
Figure 3. NDVI frequency distributions obtained during the two monitoring campaigns.  
SAVI's temporal dynamics have shown behavior like the general trend in the study but provided  
additional information at an early stage of planting. Due to the existence of soil correction in SAVI,  
this indicator is less affected by the presence of exposed soil compared to NDVI in case of partial  
canopy cover. Such property becomes especially relevant for green manure crops at the stage of  
establishment, since there is still a significant part of exposed soil in this case between each plant.  
Therefore, it is expected that SAVI will provide a more correct estimate of vegetation vigor  
compared to NDVI.  
According to the descriptive statistics, SAVI showed increasing trend from March to May, which  
is related to decreasing amount of exposed soil and increasing the area of crop canopy coverage.  
Like NDVI, the median and the upper quartile values moved to larger values, and relative  
variability decreased due to the better uniformity of distribution of vegetation on the field. The  
described trends indicate that in addition to growth of vegetation biomass, increase of vegetation  
vigor was caused by occupying exposed areas of the field by vegetation.  
879  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
Figure 4. SAVI boxplots corresponding to the March and May UAV surveys.  
The SAVI histograms validate these observations by showing a gradual movement of frequency  
of pixels towards the higher values of vegetation index in the second cycle of monitoring. In  
comparison to NDVI, March SAVI had lower occurrences of extremely low values due to the  
inclusion of soil adjustment factor that lessens the influence of soil spectrum. This makes SAVI  
more sensitive than NDVI when used to measure the development process of annual crops.  
However, in May, the distributions were similar, implying that the influence of soil exposure was  
reduced as the canopies closed.  
Figure 5. SAVI frequency distributions corresponding to March and May.  
From the joint interpretation of NDVI and SAVI, the process of vegetation growth went ahead  
according to a unified temporal pattern, with an increase in biomass, closure of the canopy, and  
decreased spatial heterogeneity. The similarity of results from both indices shows successful crop  
establishment despite mild nutritional stress revealed during laboratory analysis. Initially, areas  
with smaller vegetation index values became like others in terms of field response, suggesting  
that good physical soil properties allowed for a uniform establishment of green manure mixture.  
880  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
This conclusion is also supported by the laboratory characterization presented in the earlier  
section. The absence of physical soil limitations, as well as sufficient potassium content in the  
soils, might promote rapid growth of vegetation, while some variability in the contents of organic  
carbon and phosphorus could explain differences between field sites. Therefore, the temporal  
dynamics revealed by NDVI and SAVI represent the reaction of crops to both favorable soil  
structure and variable nutrients' availability in the experimental field.  
However, NDVI and SAVI cannot give any information on the physiological state of plants. To get  
insight into the dynamics of chlorophyll content and nutritional condition of the canopy, the  
following section will analyze behavior of the Normalized Difference Red Edge Index (NDRE). This  
index is particularly sensitive to the variations in leaf chlorophyll content and physiological  
condition of plants.  
3.4 Physiological response of the green manure crop in relation to NDRE  
Whereas NDVI and SAVI provide descriptions of the temporal evolution of canopy development  
and biomass growth, NDRE can give a different aspect of the physiological state of the green  
manure crop. In contrast to traditional vegetation indices that use the red reflectance band, the  
NDRE includes the red edge band, which is especially sensitive to the variations in the chlorophyll  
content and biochemistry of leaves. Thus, this index allows evaluating the physiological state of  
vegetation with higher sensitivity during the intermediate and late phases of plant development  
when NDVI loses its sensitivity to more biomass due to canopy closure.  
The results of descriptive statistics taken from descriptive statistics showed that NDRE had much  
lower variability compared to NDVI and SAVI indices. Although there were slight changes between  
the two campaigns, the overall variability was quite steady, showing a gradual development of  
the physiological state of vegetation while the canopy development occurred faster. This is in line  
with the natural course of development of the green manure mixtures where the fast vegetation  
coverage coincides with moderate changes in chlorophyll content.  
Figure 6. NDRE boxplots correspond to the March and May UAV surveys.  
881  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
This tendency is clearly shown by the box-and-whiskers plot of NDRE, which shows that there  
were no significant changes in median and interquartile range between different monitoring  
dates. Unlike in NDVI, in which the biomass growth caused noticeable movement of the  
distribution upwards, NDRE changed little, suggesting that the amount of chlorophyll was evenly  
distributed on the experimental plot. This means that, after the seedlings became established,  
their physiological performance went ahead in a regular way regardless of soil variability.  
Both surveys’ interquartile ranges showed very minor variations, which suggests no momentous  
changes in variability of chlorophyll concentrations over the observation period. The lack of  
momentous change in upper quartiles suggests that there were no exaggerated physiological  
responses of the most vigorous sections compared to other sections of the experimental area. On  
the contrary, the development of the crop was rather consistent and even spatially in line with  
the homogenization of canopies that occurred according to the NDVI and SAVI analysis.  
The histograms also provide some added confirmation of the above-stated conclusions. In March,  
the pixel frequency was evenly distributed on quite a narrow range around the median NDRE  
value, which is associated with the early physiological state of the crop after its establishment. In  
May, the histogram was more concentrated but kept the same central tendency. It is possible to  
assume that there were no momentous changes in chlorophyll concentration of the experimental  
field in spite of a considerable increase in biomass measured by NDVI and SAVI.  
Figure 8. NDRE frequency distributions correspond to the March and May monitoring campaigns.  
One of the features of NDRE response is the change in the symmetry of the distribution between  
different monitoring dates. Descriptive statistics show that the values of the skewness parameter  
became closer to symmetrical distribution in the case of the second UAV survey, which shows  
that the proportion of areas with reduced chlorophyll concentration decreased. In terms of  
agriculture, this trend means that the physiological heterogeneity of crops decreased with time,  
and almost all plants achieved the same physiological state by the end of the observation period.  
The stable NDRE response cannot be regarded as the lack of development of crops because it  
means that the increase in biomass and canopy occurred faster than changes in chlorophyll  
concentration. Such a complementarity of responses explains the higher temporal differences in  
882  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
the behavior of NDVI and SAVI compared to NDRE because while NDVI and SAVI mainly accounted  
for changes in plant structures caused by growing vegetation cover, NDRE describes the stable  
photosynthetic apparatus. Thus, the three vegetation indices account for different but  
complementary aspects of crop development.  
The physiological stability described by the NDRE response is also in line with the results of the  
laboratory study provided in Section 3.1. The lack of salinity stress and favorable physical  
characteristics of soils meant that no physiological stress appeared during the establishment of  
crops. Even though there were limitations due to the concentration of organic carbon and  
phosphorus, enough potassium and soil structure may support a constant level of chlorophyll  
concentration and photosynthesis.  
With respect to precision farming, the combination of NDRE significantly improves crop  
monitoring because it provides physiological data which cannot be obtained based on biomass  
indices alone. Although NDVI and SAVI show areas where vegetation is growing, NDRE helps to  
decide if the plant material has the necessary physiological state. The interpretation of these  
indices minimizes the ambiguity which appears while estimating crop development using one  
spectral index and creates a solid foundation for defining the management zones requiring  
further investigation or different crop management practices.  
Overall, the analysis of NDRE reveals that the physiology of the green manure mixture was rather  
uniform during the period of establishment of the green manure mixture, although there was  
some heterogeneity in the initial soil samples. Thus, this information, together with the structure  
data provided by NDVI and SAVI, confirms that the development of crops took place not only by  
accumulating biomass but also keeping the stable physiological status of most of the test field.  
Thus, the combination of the presented spectral responses allows making a comprehensive  
description of crop establishment in the Andean environment.  
The combined interpretation of NDVI, SAVI and NDRE allows making the assessment of canopy  
development and its physiological state. However, the spectral information from the  
multispectral images should be supported by visual data. Therefore, the next section combines  
the RGB Ortho mosaics with the multispectral indices to assess the spatial dynamics of canopy  
coverage.  
3.5 Interpretation of RGB Ortho mosaic images together with multispectral vegetation indices  
The dynamic characteristics seen through the multispectral vegetation indices have been  
confirmed visually by the RGB Ortho mosaic images collected during the two UAV surveys. While  
the RGB Ortho mosaic images cannot provide any quantification of physiological variables or  
biomass production, they give the vital visual evidence needed to confirm the spatial patterns  
shown from the multispectral analysis. In simple terms, the use of RGB Ortho mosaic images  
alongside the NDVI, NDRE, and SAVI vegetation indices allows for a better understanding of crop  
growth by using both visual and spectral analyses.  
883  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
Based on the Ortho mosaic image collected during the first survey; it can be noted that the crops  
were still in the establishment stage in March. There was no uniformity in vegetation coverage  
since there were many patches of bare soil present amidst the developing crops. This spatial  
heterogeneity is consistent with the statistical features of NDVI and SAVI, which show low  
medians and high interquartile because of the large variation in the vegetation coverage in March.  
Figure 9. RGB Ortho mosaics acquired.  
Secondly, it should be mentioned that there were clear differences in the visual aspect of the  
Ortho mosaics produced during the second period of monitoring. By May, the field was covered  
by more vegetation, hence, the higher density of the green canopy with fewer bare spots. This  
enhanced homogeneity of the green canopy is represented by a shift in the graphs of NDVI and  
SAVI towards higher values, hence, suggesting an overall growth for biomass rather than just the  
patches of vigorous vegetation. Therefore, the use of RGB Ortho mosaics can serve as visual  
evidence of the increased canopy closure due to biomass growth inferred from multispectral  
indices.  
Figure 10. RGB Ortho mosaics obtained.  
Also, descriptive statistics for RGB Ortho mosaics support the conclusion above. First, the  
interquartile range in the values of spectral response in visible channels became lower during the  
second period of monitoring, thus, suggesting that vegetation development had resulted in higher  
884  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
homogeneity of the visible signal. Secondly, the change in the histograms suggests that the  
variability of the surfaces with different reflectance properties (i.e., soil, crops, and young  
vegetation) has decreased. This conclusion is completely consistent with the changes in the values  
of NDVI and SAVI mentioned above.  
Another benefit of using RGB Ortho mosaics is the potential of using this type of data in terms of  
easing interpretation of the spatial patterns of plant growth. Unlike vegetation indices that only  
provide a numerical value of the spectral response, RGB imagery can directly detect the spatial  
distribution of dense vegetation cover, sparse canopy, borders of the fields, and heterogeneous  
areas. Such visual information is especially helpful in interpretation of the multispectral product  
since this will help correlate any spectral anomalies with features in the field rather than interpret  
the statistical distribution only. In other words, RGB imagery adds more to the existing analysis of  
multispectral indices rather than substitutes it.  
Combining RGB imagery with laboratory analysis of the soils reveals even more about the  
processes controlling the crop development. According to laboratory tests, soil physical  
properties were favorable for root growth and water flow, while variability in organic carbon and  
phosphorus was the main nutrient factor limiting crop growth. According to the RGB Ortho  
mosaic images, however, this nutrient variability did not prevent the crops from being proved on  
this field; instead, it may have affected the rate of canopy closure in some sectors. In March, when  
plants were showing, there were areas that developed their canopy slower than others; in May,  
this difference became much less pronounced.  
Integrated interpretation of RGB imagery and the analysis of vegetation indexes illustrates the  
benefits of including other data in UAV-based scouting of crops. NDVI and SAVI measure  
structural changes associated with plant growth, while NDRE measures the physiological state of  
vegetation. Conversely, RGB Ortho mosaics offer a necessary context to ensure that any change  
in the values of vegetation indexes is caused by an actual change in the plant canopy and not by  
any spectral anomalies or errors in the image processing. A high degree of consistency across  
various data sources makes the interpretation of the process of crop growth much more  
dependable.  
From a practical standpoint, the use of RGB and multispectral images serves as an example of one  
of the key benefits offered by the UAVs. A single flight of the UAV allowed obtaining regular Ortho  
mosaics suitable for visual analysis of the field as well as for obtaining multispectral information  
useful for assessing the physiological state of plants.  
The combination of soil analysis in the laboratory, RGB Ortho mosaic images, and vegetation index  
analysis confirms that the green manure mixture was proved taking into consideration the  
physicochemical properties of the field. Positive canopy closure, increased plant vitality, and good  
physiological state confirm that the crop was positively affected by soil conditions in the field  
despite the somewhat uneven distribution of nutrients. It is worth noting that consistency  
between visual evaluation, descriptive statistics, and vegetation index analysis supports the  
validity of the methods used during the current study.  
885  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
To conclude, the vegetation indices generated using the UAV imagery cannot be considered  
separately. A more holistic picture of crop growth appears when one analyzes laboratory soil test  
results, RGB Ortho mosaic images, and added vegetation indices. The results obtained have  
significance for further discussion based on earlier studies.  
DISCUSIÓN  
Combining soil analysis in the laboratory and remote sensing data collected using UAVs with  
multispectral cameras provided a complete understanding of the main parameters affecting the  
formation and early growth of the green manure mixture. At the same time, analyzing soil  
properties and spectral properties of the plants rather than only vegetation indices enabled  
naming the main parameters of plant growth during the monitoring period.  
The laboratory soil analyses showed that the studied soil had good physical properties for plant  
cultivation because of low bulk density, low salinity level, and soil texture being sandy loam and  
sandy clay loam. In such soil, plants grow well because of proper soil aeration, no obstacles to  
water percolation, and easy penetration by roots. However, low levels of organic carbon and  
phosphorus made the soil nutrients quite heterogeneous throughout the experimental area. Such  
heterogeneity of soil nutrient content is likely to contribute to the spatial heterogeneity of plant  
spectral response at the initial stage of UAV monitoring when vegetation with low NDVI and SAVI  
values did not develop properly yet.  
The contrasting behavior observed between NDVI and NDRE can be explained by the physiological  
sensitivity of each vegetation index during crop development. NDVI responded markedly to the  
increase in green biomass and canopy closure from March to May, reflecting the progressive  
accumulation of photosynthetically active vegetation. Similar responses have been reported in  
precision agriculture studies, where NDVI is considered highly effective for monitoring biomass  
accumulation during the early and intermediate growth stages of crops (3,4). In contrast, NDRE  
remained comparatively stable despite the substantial increase in biomass. This behavior is  
consistent with previous studies indicating that NDRE is more closely associated with chlorophyll  
concentration than with total biomass and becomes particularly useful after canopy closure,  
when NDVI tends to approach saturation under dense vegetation conditions (4,23). The relatively  
stable NDRE values observed in this study therefore suggest that, although biomass increased  
considerably during crop establishment, the physiological status and chlorophyll concentration of  
the green manure canopy remained relatively uniform across the experimental area. Similar  
responses have been reported for agricultural systems under high-Andean environmental  
conditions, where NDRE has proven to be a reliable indicator of plant physiological status despite  
significant structural changes in canopy development (22,24,25). Consequently, the  
complementary use of NDVI and NDRE provides a more comprehensive assessment of crop  
development, combining information on biomass production with physiological condition and  
reducing the limitations associated with the use of a single vegetation index.  
On the contrary, changes of NDRE are quite small during the examined period despite the obvious  
increases of the vegetation biomass. It corresponds to the fact that red-edge region is  
886  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
physiologically sensitive to the chlorophyll concentration in the crops and not to the total amount  
of biomass. Thus, the stability of the NDRE can be interpreted as a relatively stable content of  
chlorophyll during the considered growth stage, whereas the main changes occurred via canopy  
closure and increasing vegetation coverage. The complementarity of the responses proves the  
effectiveness of using various vegetation indices for analysis of crop growth.  
Although statistically significant correlations were identified for selected variables, the limited  
number of soil sampling points suggests that these relationships should be interpreted with  
caution. Nevertheless, the correlation analysis reinforces the descriptive patterns observed in the  
vegetation indices and highlights the importance of integrating soil physicochemical information  
with UAV-derived spectral data for crop monitoring under high-Andean conditions. Future studies  
including a larger number of sampling sites and repeated monitoring campaigns are  
recommended to validate these relationships.  
It was especially helpful at the establishment stage since it reduced the impact of the exposed  
soil on the spectra, allowing vegetation to be seen more easily when there is an incomplete  
canopy closure. In the first survey, soil background had a significant contribution to the variability  
in the spectral characteristics while, in the second one, there was a gradual decrease of this  
influence due to the increase in vegetation cover. Therefore, the use of NDVI and SAVI together  
provided a more precise assessment of crop development than the separate use of any of these  
indexes, particularly in those areas where there is no complete closure of the canopy.  
The RGB Ortho mosaic images could be considered a qualitative proof of the temporal trends  
found in multispectral analyses. Though the RGB images do not provide quantitative  
measurements of vegetation health, the reduction of exposed soil and the increase in the  
greenness of the canopy were consistent with the trends found in the spectral indexes. Thus, this  
agreement of qualitative and quantitative data proves that the remote sensing technique applied  
in this research is dependable.  
About precision agriculture, the main innovation brought about by this study is its ability to show  
that using UAV multispectral imagery allows one to precisely describe the space-time dynamics  
of showing green manure cover crops as well as the effects of soil physicochemical heterogeneity  
on their growth. It should be noted that remote sensing is not intended to replace the laboratory  
soil analysis but rather supplement it through analyzing the whole field experiment plot and  
therefore making zoning and site-specific agronomy possible. Hence, the use of UAV imagery  
along with GIS and soil laboratory data seems like an effective way to check cover crops and  
implement sustainable soil management in Andean agriculture.  
Furthermore, this research proves the effectiveness of UAV monitoring during the establishing  
period. It would be helpful to have more information on biomass, chlorophyll content, nutrients  
absorption, and Mult seasonal studies since this will allow one to interpret the spectral response  
of the crop under different environmental conditions. In addition, the use of machine learning  
techniques applied to multispectral imagery and soil data can improve precision agriculture  
algorithms.  
887  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
CONCLUSIONES  
From the results shown above, it is evident that the use of soil physicochemical analysis in the  
lab, multispectral images captured by the use of an unmanned aerial vehicle and GIS form a  
powerful framework when used together in assessing the development of green manure crops in  
the Andean environment.  
From the preliminary results obtained from the laboratory analysis of the soil, it was clear that  
the physical characteristics of the soil in the experimental site were favorable for the  
establishment of crops since they had low salinity levels, good soil structure and texture. While  
organic carbon and available phosphorus in the soil were relatively low and hence could be  
considered as some nutritional constraints to the growth of crops, it did not hinder the successful  
establishment of crops.  
Based on the above findings, it is clear that a combination of laboratory soil physicochemical  
characterization, multispectral UAV imaging, and geographic information system is an efficient  
way to track the growth of the green manure crop in the Andean environment.  
With respect to the laboratory soil characterization, it was found that the field had favorable  
physical properties conducive to plant growth, based on low levels of salinity, proper soil  
structure, and proper soil texture. However, organic carbon and available phosphorus were  
relatively low, possibly inhibiting nutrient availability.  
REFERENCIAS BIBLIOGRÁFICAS  
1.  
Rouse JW Jr, Haas RH, Schell JA, Deering DW. Monitoring vegetation systems in the Great  
Plains with ERTS. In: Third Earth Resources Technology Satellite-1 Symposium. Washington  
(DC): NASA; 1974. p. 309-317.  
2.  
3.  
4.  
5.  
6.  
7.  
Huete AR. A soil-adjusted vegetation index (SAVI). Remote Sens Environ. 1988;25(3):295-  
309. doi:10.1016/0034-4257(88)90106-X.  
Tucker CJ. Red and photographic infrared linear combinations for monitoring vegetation.  
Remote Sens Environ. 1979;8(2):127-150.  
Hunt ER Jr, Daughtry CST. What good are unmanned aircraft systems for agricultural remote  
sensing and precision agriculture? Int J Remote Sens. 2018;39(15-16):5345-5376.  
Zhang C, Kovacs JM. The application of small unmanned aerial systems for precision  
agriculture: A review. Precis Agric. 2012;13(6):693-712. doi:10.1007/s11119-012-9274-5.  
Colomina I, Molina P. Unmanned aerial systems for photogrammetry and remote sensing:  
A review. ISPRS J Photogramm Remote Sens. 2014;92:79-97.  
Berni JAJ, Zarco-Tejada PJ, Suarez L, Fereres E. Thermal and narrowband multispectral  
remote sensing for vegetation monitoring from an unmanned aerial vehicle. IEEE Trans  
Geosci Remote Sens. 2009;47(3):722-738.  
8.  
Mulla DJ. Twenty-five years of remote sensing in precision agriculture: Key advances and  
remaining knowledge gaps. Biosyst Eng. 2013;114(4):358-371.  
888  
EVALUACIÓN ESPACIOTEMPORAL DEL CRECIMIENTO DEL ABONO VERDE MEDIANTE TELEDETECCIÓN MULTIESPECTRAL CON  
UAV, SIG Y ANÁLISIS FISICOQUÍMICO DEL SUELO  
9.  
Zhang H, Wang Y. A review of the application of UAV multispectral remote sensing  
technology in precision agriculture. Smart Agric Technol. 2025;12:101406.  
Grbović Ž, Ivošević B, Budjen M, Waqar R, Pajević N, Ljubičić N, et al. Integrating UAV  
multispectral imaging and proximal sensing for high-precision cereal crop monitoring. PLoS  
One. 2025;20(5):e0322712.  
10.  
11.  
12.  
13.  
14.  
FAO. Conservation Agriculture. Rome: Food and Agriculture Organization of the United  
Nations; 2021.  
Lal R. Soil carbon sequestration impacts on global climate change and food security. Science.  
2004;304(5677):1623-1627.  
Brady NC, Weil RR. The Nature and Properties of Soils. 15th ed. Upper Saddle River (NJ):  
Pearson; 2017.  
Doran JW, Parkin TB. Defining and assessing soil quality. In: Doran JW, Coleman DC, Bezdicek  
DF, Stewart BA, editors. Defining Soil Quality for a Sustainable Environment. Madison (WI):  
SSSA; 1994. p. 3-21.  
15.  
Blanco-Canqui H, Lal R. Principles of Soil Conservation and Management. Dordrecht:  
Springer; 2008.  
16.  
17.  
Fageria NK. The Use of Nutrients in Crop Plants. Boca Raton (FL): CRC Press; 2009.  
Precision Agriculture Committee. Precision agriculture: Definition and concepts. Madison  
(WI): ASA; 2019.  
18.  
19.  
20.  
21.  
22.  
Gebbers R, Adamchuk VI. Precision  
2010;327(5967):828-831.  
McBratney A, Whelan B, Ancev T, Bouma J. Future directions of precision agriculture. Precis  
Agric. 2005;6:7-23.  
Bongiovanni R, Lowenberg-Deboer J. Precision agriculture and sustainability. Precis Agric.  
2004;5:359-387.  
Cambardella CA, Karlen DL. Spatial analysis of soil fertility parameters. Precis Agric.  
1999;1:5-14.  
Leta Morocho PV. Correlación espacial de índice de vegetación de diferencia normalizada  
(NDVI) entre plataformas satelitales y vehículos aéreos no tripulados (UAV), en los  
agroecosistemas de cuatro cuencas de los altos Andes del macizo del Cajas (tesis). Cuenca:  
Universidad Católica de Cuenca; 2023.  
agriculture  
and food security. Science.  
23.  
24.  
25.  
Vera-Esmeraldas A, et al. UAV-Based Spectral and Thermal Indices in Precision Agriculture.  
Agronomy. 2025;15(11):2569.  
Jara C, et al. Study of wetlands in the Ecuadorian Andes through comparison of Landsat-8  
and Sentinel-2 imagery. Revista de Teledetección. 2019.  
Zea P, et al. NDVI Performance for Monitoring Agricultural Energy Inputs in the Ecuadorian  
Andean Region. 2025.  
889