Spatiotemporal Dynamics of Land Use and Land Cover through Physical–Hydraulic Indices: Insights in the São Francisco River Transboundary Region, Brazilian Semiarid Area

Abstract

This article presents a study on the spatiotemporal dynamics of land cover and use, vegetation indices, and water content in the semiarid region of Pernambuco, Brazil. This study is based on an analysis of satellite images from the years 2016, 2018, and 2019 using the MapBiomas platform. The results show changes in the predominant land cover classes over time, with an increase in the caatinga area and a decrease in the pasture area. An analysis of the vegetation indices (NDVI and LAI) indicated low vegetation cover and biomass in the study area, with a slight increase in the NDVI in 2018. An analysis of the Modified Normalized Difference Water Index (MNDWI) showed that the water content in the study area was generally low, with no significant variations over time. An increase in the water bodies, mainly due to the construction of a reservoir, was noted. The results of this study have provided important information for natural resource management in the region, including the development of strategies for the sustainable use and management of natural resources, particularly water resources, vegetation cover, and soil conservation.

1. Introduction

Semiarid regions, despite being subject to intense rainfall events, have temporally and spatially irregular rains, which cause extreme situations such as droughts and floods [1,2,3,4]. In addition, these regions have high evaporation rates and shallow soils with low water retention capacities, which increases water scarcity problems [5,6,7]. Such characteristics make these regions susceptible to desertification. In the Brazilian semiarid region, there is the only exclusively Brazilian biome, the caatinga, which is predominantly present in the northeast region and has the characteristic of losing its leaves during the dry season, in addition to other attributes that provide the biome with high resistance to drought [4,8,9].

The northeast region is among the areas susceptible to the desertification process, with semiarid and sub-humid climates and subject to intense annual rainfall variations and vegetation suppression [10,11,12,13,14]. Removal of the vegetation cover causes soil erosion, which reduces the water retention capacity and consequently decreases the soil organic matter, leading to soil degradation, threatening biodiversity, and reducing plant biomass [15]. Furthermore, changes in land use and occupation are expected to occur in the semiarid region of the state of Pernambuco, either due to natural causes related to climate change or induced by the implementation of the São Francisco River Integration Project (PISF), also referred to as the São Francisco River Transboundary Project, with the hydrographic basins of the northern northeast [16,17]. The axes, branches, and pipelines were designed for human and animal water supplies and should also contribute to the expansion of the water supply in Pernambuco [18,19].

The environmental conditions of the semiarid region strongly influence the land use dynamics, as well as biophysical parameters such as the albedo, vegetation, and water indices [20,21]. Environmental monitoring carried out directly in the field is often unfeasible due to operational cost, rainfall seasonality, and the spatial variability of the soil and vegetation characteristics [8]. Thus, the use of remote sensing techniques and a geographic information system (GIS) has been increasingly encouraged for semiarid conditions [22]. The use of remote sensing to obtain data related to land use and the biophysical parameters of the environment in order to conduct studies and monitor natural resources, soil degradation, and vegetation cover has increased in recent years and can assist in the adoption of sustainable practices and management [15,21,23,24].

Studies utilizing very high-resolution land cover data and phenological metrics derived from remote sensing have been instrumental in assessing vegetation status and changes [25,26,27,28]. For example, Orusa et al. [25] developed an Earth observation service to map land cover in complex geomorphological areas, enabling continuous mapping with high spatial and temporal resolutions. Another study by Pielke et al. [26] highlighted the impacts of human land management on landscape structure and surface fluxes, emphasizing the need to consider these changes in climate assessments and adaptation strategies. Additionally, Orusa et al. [27] developed an algorithm for mapping phenological metrics in mountain areas worldwide, utilizing high-resolution satellite data to gain insights into land surface phenology and trends.

Biophysical indices (NDVI, LAI, and MNDWI) are available through satellite image processing and have been important input for identifying land cover and land use changes. From temporal analysis of VIs, it is possible to define patterns that characterize a particular culture or group of cultures or even a particular land use dynamic [29]. By monitoring the water behavior associated with vegetation cover variation, it is possible to diagnose anthropogenic and natural alterations in the landscape and guide the establishment of policies to reverse an environmental degradation scenario [30,31,32]. The Modified Normalized Difference Water Index (MNDWI) developed by Xu [33] has been widely accepted for identifying surface water bodies. This index has proven to be efficient for mapping small reservoirs in hydrographic basins in the semiarid region of Pernambuco [32].

The association of vegetation indices, such as the NDVI and LAI, with such water indices as the MNDWI, using remote sensing techniques and geoprocessing of orbital images, is necessary to assess the intensification of anthropogenic processes. In a study by Silva et al. [34], which evaluated the spatiotemporal changes in environmental degradation processes of vegetation cover and water resources through physical and water-related parameters (NDVI and MNDWI) on the surface, using remote sensing techniques, in northeastern Brazil, the authors observed the impacts of the degradative processes of anthropogenic actions and their intensification on water bodies and vegetation dynamics. Consistently with the aforementioned study, Silva et al. [35], who explored the NDVI, LAI, and MNDWI to characterize natural and anthropogenic degradation processes in a microregion of northeastern Brazil, also emphasized in their study results the importance of applying physical and water indices to highlight soil use and land cover degradation processes. Furthermore, this study addresses the need to fill existing knowledge gaps regarding the environmental and agricultural impacts of the São Francisco River Transboundary Project and some of its associated reservoirs.

Therefore, the present study aims to analyze the land cover and land use changes, vegetation dynamics, and water bodies in a specific region located between the municipalities of Terra Nova and Cabrobó in the state of Pernambuco, Brazil. This study sought to assess the impacts of agricultural activities and the presence of reservoirs, particularly those associated with the São Francisco River Transboundary Project, on the region’s environment. Furthermore, this study examines spatiotemporal variations in vegetation indices such as the Normalized Difference Vegetation Index (NDVI), the Leaf Area Index (LAI), and the Modified Normalized Difference Water Index (MNDWI) to gain insights into vegetation vigor, canopy structure, and water body dynamics. The findings of this study have the potential to contribute significantly to decision-making processes related to natural resource management and conservation in the study area.

2. Materials and Methods

2.1. Characterization of the Study Site

The study area, depicted in Figure 1, is situated between the municipalities of Terra Nova and Cabrobó, within the geographic coordinates of 08°17′38″ S and 39°24′14″ W and of 08°10′29″ S and 39°17′31″ W. It encompasses a total area of 16,481.3 hectares and falls within the São Francisco Mesoregion and the Petrolina Microregion in the state of Pernambuco [36,37,38,39]. The study area intersects with the north axis of the São Francisco River Integration Project (PISF), also referred to as the São Francisco River Transboundary Project. The São Francisco River Transboundary Project is a significant water initiative led by the Federal Government of Brazil, spanning a length of 477 km. It comprises two primary axes, the north axis and the east axis, incorporating four tunnels, 14 aqueducts, nine pumping stations, and 27 reservoirs, with 18 of these reservoirs situated in Pernambuco. This project’s objective is to secure water supplies for the socio-economic development of drought-prone states such as Ceará, Paraíba, Pernambuco, and Rio Grande do Norte [18,40,41].

According to the Köppen–Geiger climate classification, the climate of the region is type BSh (hot, semiarid climate) [42,43], with an average annual rainfall of approximately 600 mm. The wet season begins in January and extends until April [44]. The predominant soil classes are Entisols, Entisols (Fluvents), Alfisols (Natrustalfs and Natrudalfs) and Alfisols and Aridisols (Argids). The vegetation is composed of hyperxerophilic caatinga, with sections of deciduous forests. The relief is predominantly smooth-undulation. Both municipalities are located in the geoenvironmental unit of the Sertaneja Depression and, geologically, in the Borborema Province [36,37,45].

Three main water bodies were identified in the analyzed region. The first is the Nilo Coelho reservoir, located between the municipalities of Terra Nova and Cabrobó. The second and third are the Terra Nova and Serra do Livramento reservoirs, located in Cabrobó, to the north of the municipality. The Nilo Coelho reservoir is used as a water supply source for human consumption and for irrigation use in the municipality, being one of the largest in the Sertão of Pernambuco, with a capacity of 22 million m³, and was constructed in 1928 [46]. The Terra Nova and Serra do Livramento dams were constructed as part of the north axis of the São Francisco River Transboundary Project, and from the opening of their gates, the waters of the Transboundary Project can be released for human and animal consumption, as well as to increase water supply for irrigation and stream perenization [18,19,47].

2.2. Remote Sensing Data

Three Landsat-8 satellite images were selected; they were acquired from the website https://earthexplorer.usgs.gov/ (accessed on 4 June 2023), an American spatial database of the National Aeronautics and Space Administration/United States Geological Survey (NASA/USGS). The Landsat-8 Operational Land Imager/Thermal Infrared Sensor (OLI/TIRS) is composed of 11 multispectral bands ranging from 0.43 to 12.51 µm in wavelength, with spatial resolutions of 30 m (bands 1 to 9), except for bands 10 and 11, which have spatial resolutions of 100 m; spectral resolutions of 16 bits, representing gray levels from 0 to 65,535; and a temporal resolution of 16 days.

The images used corresponded to the dry period of the region and to the dates of 29 October 2016, 17 September 2018, and 15 October 2019, with orbit 217 and point 66, except for the year 2018, which had orbit 216 and point 66, as shown in

Table 1. Image date, time, sun elevation angle, and satellite point.

Date	UTM Time	Sun Elevation Angle	Orbit	Point
29 October 2016	12:48:21	65.9414	217	66
17 September 2018	12:47:38	61.1147	216	66
15 October 2019	12:42:13	65.7611	217	66

Source: USGS [48].

. All images from 2017 for the study area presented clouds; therefore, no analysis was carried out for that year.

2.3. Rainfall Data

Rainfall data for the years of the imaging dates were analyzed for the study area from 2016 to 2019 using data obtained from the Pernambuco Agency of Water and Climate (APAC)’s Terra Nova weather station and the National Institute of Meteorology (INMET)’s Cabrobó automatic weather station, as shown in Figure 2. The average annual total rainfalls were 490.55 mm (2016), 533.1 mm (2018), and 417.2 mm (2019).

2.4. Land Use and Land Cover Classifications

The land use and land cover classifications for Brazil were based on images freely available on the online platform MapBiomas, which has classifications, available up to 2019, based on Landsat satellite images with spatial resolutions of 30 m. Maps were generated for the years 2016, 2018, and 2019. The classified images were from Collection 5, which uses a catalog with 27 classes [50]. Afterward, the area of the class related to water bodies was calculated to quantify the evolution of these areas over time.

2.5. Calculation of Water and Vegetation Indices

The Normalized Difference Vegetation Index (NDVI) is the most commonly used vegetation index and is an indicator of photosynthetically active biomass. It is calculated as the ratio of the differences between the near-infrared (${\mathsf{\rho }}_{NIR}$) and red (${\mathsf{\rho }}_R$) bands, respectively to their sums, as shown in Equation (1) [51]:

$$\mathrm{NDVI}=\frac{{\mathsf{\rho }}_{NIR} - {\mathsf{\rho }}_R}{{\mathsf{\rho }}_{NIR}+{\mathsf{\rho }}_R}$$

(1)

The Soil-Adjusted Vegetation Index (SAVI) was calculated according to Equation (2), and it is an indicator of the amount and condition of green vegetation that seeks to mitigate the background effects of soil [52]:

$$\mathrm{SAVI}=\frac{\left(1 + \mathrm{L}\right)\left({\mathsf{\rho }}_{NIR} - { \mathsf{\rho }}_R\right)}{\mathrm{L}+{\mathsf{\rho }}_{NIR}+{ \mathsf{\rho }}_R}$$

(2)

where ${\mathsf{\rho }}_R$ and ${\mathsf{\rho }}_{NIR}$ are the red and near-infrared bands, respectively, and L is a factor depending on the type of soil, considered to be 0.5 for this study [7,20,53,54,55].

The Leaf Area Index (LAI) is defined as the ratio of the total leaf area to the area of land used by that vegetation. It is an indicator of the biomass of each pixel of an image, computed with Equation (3) [53]:

$$\mathrm{LAI}=\frac{-\ln \left(\frac{0.69 - \mathrm{SAVI}}{0.59}\right)}{0.91}$$

(3)

The Modified Normalized Difference Water Index (MNDWI) was proposed by Xu [33] and aims to highlight water features and identify watercourses. It is calculated as the ratio between the differences of the reflectance values of the visible green (${\mathsf{\rho }}_{\mathrm{G}}$) and medium infrared (${\mathsf{\rho }}_{\mathrm{M}\mathrm{IR}}$) bands, divided by their respective sums, as shown in Equation (4). The MNWDI ranges from −1 to +1, and values above zero are indicative of the presence of water bodies. Subsequently, the delimitation of the water bodies was vectorized and exported and their area calculated.

$$\mathrm{MNDWI}=\frac{{\mathsf{\rho }}_{\mathrm{G}} -{ \mathsf{\rho }}_{\mathrm{M}\mathrm{IR}}}{{\mathsf{\rho }}_{\mathrm{G}}+{\mathsf{\rho }}_{\mathrm{M}\mathrm{IR}}}$$

(4)

2.6. Image Processing and Statistical Analysis

All images were processed in QGIS software version 3.22, and reflective bands from 2 to 7 were used in the Landsat-8 satellite images. To determine the vegetation and water indices, the Semiautomatic Classification Plus (SCP) plugin was used, converting gray levels to radiance and reflectance using radiometric calibration coefficients found in the image metadata. Subsequently, processing was carried out considering the scene cuts according to the areas of interest, and the images were reprojected to SIRGAS2000.

For the visualization of the region, false color composites of bands 5, 4, and 3 (for the Landsat-8 images) were used for the near-infrared, red, and green bands, respectively, as shown in Figure 3.

The QGIS “Raster Calculator” tool was used to calculate the NDVI, LAI, and MNDWI. Afterward, the values of these indices for all pixels within the region were extracted. These extracted values were then subjected to descriptive statistics, including measures such as the minimum, maximum, mean, standard deviation (SD), and coefficient of variation (CV).

3. Results

The annual thematic classification maps of land cover and use for the study area in the years 2016, 2018, and 2019 were developed based on the MapBiomas platform, which allowed us to infer the main cover classes that make up the site (Figure 4 and

Table 2. Area (hectares) and percentages (%) of land use and land cover classifications.

Land Use and Land Cover Classification	2016 (ha)	2016 (%)	2018 (ha)	2018 (%)	2019 (ha)	2019 (%)
Forest Formation	7.07	0.04%	15.74	0.10%	25.56	0.16%
Savanna Formation	7053.07	43.68%	7237.80	44.83%	7503.64	46.48%
Grassland Formation	3232.31	20.02%	3759.63	23.29%	3725.85	23.08%
Pasture	2495.44	15.46%	3020.41	18.71%	2974.25	18.42%
Mosaic of Agriculture and Pasture	3269.79	20.25%	1382.94	8.57%	1327.67	8.22%
Urban Infrastructure	29.89	0.19%	30.42	0.19%	30.33	0.19%
Water Bodies	57.92	0.36%	698.56	4.33%	558.21	3.46%

). The year 2016 (Figure 4a) presented coverage percentages of forest, savanna, and grassland formations of 0.04, 43.68, and 20.02%, respectively. In 2018 (Figure 4b), there were increases, with values of 0.10, 44.83, and 23.29%, respectively. Similarly, for 2019 (Figure 4c), the percentages were 0.16, 46.48, and 23.08%, respectively.

Thus, the transition from 2016 (Figure 4a) to 2018 (Figure 4b) showed an increase in the caatinga area (forest, savanna, and grassland formations) of 7.00% and a reduction in the agriculture and pasture areas (pastures and agriculture–pasture mosaic) of 23.62%. When 2016 and 2019 were compared (Figure 4c), there were a 9.35% increase in the caatinga area and a 25.38% reduction in the agriculture and pasture areas. This indicates a shift in land use practices and a potential conversion of agricultural and pasture areas into natural vegetation formations.

Another factor that changed the landscape was the increase of 1106.08% in the areas covered by water bodies from 2016 to 2018, mainly due to an increase in the volume of the Nilo Coelho reservoir and the completion of the Transboundary axis, which includes the Terra Nova and Serra do Livramento reservoirs. These developments have contributed to increased water availability in the region. It is worth noting that despite being constructed in 1928, the Nilo Coelho reservoir was completely dry in the year 2016 (Figure 4a).

Figure 5 shows the spatiotemporal distributions of the Normalized Difference Vegetation Index (NDVI) for 2016, 2018, and 2019, in the dry season, where the values range from 0 to 0.6. NDVI values between 0 and 0.3 are considered more susceptible to degradation, presenting areas with bare soil and sparse vegetation, according to false-color composition images (Figure 3). The NDVI is an index related to vegetation activity, and the closer it is to 1, the higher the photosynthesis and transpiration are [56]. When the NDVI is between 0 and 0.30, it indicates the occurrence of herbaceous vegetation or that the caatinga is leafless [57].

Areas with NDVI values of greater than 0.45 represent regions of high vegetative vigor. Therefore, it can be observed in Figure 5b that these areas are more evident along the reservoir margins and near the drainage network, indicating that the soil moisture is preserved even during the dry season. However, because they have fertile soils and preserve soil moisture for a longer time, the regions close to the drainage networks have their native vegetation constantly removed [32]. The total annual precipitation in 2018 was high, which may have also influenced the result, as in 2019, the total annual precipitation was the lowest compared to other years.

Figure 6 shows the maps of the Leaf Area Index (LAI), which present values ranging from 0 to 0.5 m²/m² for the study area over the analyzed period; lower values of the LAI correspond to lower values of the Soil Adjusted Vegetation Index (SAVI). The distribution of the LAI is similar to the distribution of the NDVI; however, the LAI is a parameter that is related to the structure of the canopy, consequently affecting rain interception, light interception, and gas exchange [58]. Additionally, it is worth noting that the caatinga vegetation is highly dynamic, with accelerated growth during the wet season and leaf loss during the dry season [56]. The satellite images used in the present study were obtained during dry months; consequently, we found low LAI values due to the native vegetation losing its leaves.

Figure 7 shows the spatiotemporal distributions of the Modified Normalized Difference Water Index (MNDWI) for 2016, 2018, and 2019, where the values range from 0 to −1.0. This index minimizes information about vegetation and highlights water bodies (reservoirs, lakes, dams, and rivers) [33].

Figure 8 shows the respective surface areas of the reservoirs, determined through land use classification using the MapBiomas product and also through geoprocessing. For this, it was necessary to vectorize the regions corresponding to the water features extracted from the MNDWI on the dates of 29 October 2016, 17 September 2018, and 15 October 2019, shown in blue tones in Figure 7. Using the automated classification acquired with MapBiomas Collection 5, it was possible to observe similarities with the visual interpretation that the MNDWI maps demonstrated regarding the water bodies in the study area. As depicted in Figure 4, the image classifications for those years displayed similarities in the configurations of the water body surface area. The differences between the MNDWI and MapBiomas are related to the Landsat 8 image dates not coinciding with MapBiomas and to how the MNDWI has the potential to identify small reservoirs [59] that contribute to a larger total area of water bodies.

For a statistical analysis of the collected data,

Table 3. Statistical data of the NDVI, LAI, and MNDWI.

Date	Minimum	Maximum	Mean	SD 1	CV 2 (%)
	NDVI
29 October 2016	−0.483	0.953	0.356	0.086	24.22%
17 September 2018	−0.591	0.919	0.344	0.118	34.38%
15 October 2019	−0.684	0.870	0.288	0.121	42.09%
	LAI
29 October 2016	−0.377	5.800	0.237	0.145	60.95%
17 September 2018	−0.362	4.860	0.241	0.184	76.33%
15 October 2019	−0.380	6.330	0.190	0.170	89.65%
	MNDWI
29 October 2016	−0.902	0.492	−0.708	0.068	9.56%
17 September 2018	−0.862	0.567	−0.645	0.169	26.23%
15 October 2019	−0.734	0.641	−0.581	0.168	28.97%

¹ SD = standard deviation, ² CV = coefficient of variation.

presents the minimum, maximum, mean, standard deviation, and coefficient-of-variation values for each year. It is evident that the NDVI mean on 29 October 2016 (0.356) was higher compared to those on 17 September 2018 (0.344) and 15 October 2019 (0.288). The NDVI values serve as an indicator of vegetation vigor in the study area. The higher NDVI mean in 2016 suggests a healthier vegetation state during that period. However, subsequent years exhibited a decrease in the mean NDVI, indicating a potential decline in vegetation vigor. Several factors could contribute to this decline, including drought conditions or changes in land use, such as the construction of two reservoirs in the study area.

The MNDWI values provide valuable insights into the presence of water bodies, with negative values close to zero indicating the presence of water. It is evident that in 2019 (−0.581), the negative MNDWI values were more prominent compared to in 2016 (−0.708) and 2018 (−0.645), suggesting a higher occurrence of water bodies during that period. Furthermore, the coefficient of variation (CV) values serve as an indicator of data heterogeneity or variability. In 2019, both the NDVI (42.09%) and the MNDWI (28.97%) exhibited higher CV values compared to those of 2016 and 2018. This indicates that the data for 2019 displayed greater heterogeneity, suggesting more diverse responses of vegetation and water bodies to environmental conditions.

This increased heterogeneity in 2019 could potentially be attributed to factors such as elevated rainfall or changes in water management practices in the region, such as the implementation of the Transboundary Project. The presence of water bodies plays crucial roles in sustaining vegetation and preserving soil moisture, particularly in arid or semiarid regions like the study area. The high dispersion observed in the vegetation indices can be attributed to the various types of soil cover present in the study area, with seven distinct types identified.

4. Discussion

The results of this study suggest that there were changes in the land cover and use in the study area between the time periods of 2016–2018 and 2016–2019, including an increase in the caatinga area and reductions in agriculture and pastures. The increase in the native vegetation areas may be related to the severe drought that occurred in the region between the years 2012 and 2015; in view of these scenarios, several farmers reduced or abandoned agricultural activities, favoring the natural regeneration process of native vegetation [60]. Similar results were obtained by Silva et al. [8] in the transition between 2017 and 2018, when they studied changes caused by agricultural activity in a semiarid region with a predominant caatinga biome in a historical series from 1998 to 2018, with orbital images from Landsat-5 and Landsat-8 satellites. They found that an increase in agriculture resulted in the degradation of the caatinga biome areas and that the recession of livestock farming promoted the natural recovery of the biome.

These results also highlight the importance of reservoirs in the socio-economic development of the region, as they enable irrigation of small areas, guarantee water for livestock and human consumption, and contribute to the recharge of underground aquifers and reductions in potential flash floods [61]. Bi et al. [62] emphasized that the MNDWI is efficient for identifying water reservoirs. We could see that this method identified more reservoirs with smaller water surfaces than did the survey carried out with MapBiomas (Figure 3).

The spatiotemporal distributions of the NDVI, LAI, and MNDWI in the study area provided valuable information about the vegetation vigor, canopy structure, and water bodies. The caatinga biome, where the study area is located, experiences distinct wet and dry seasons, which significantly impacts vegetation dynamics. This natural variability in weather patterns, characterized by periods of rainfall followed by drought, may explain the observed variations in the vegetation indices across different years [56].

Additionally, it is important to consider the influence of the construction of two reservoirs in the study area through the São Francisco River Transboundary Project. The presence of these reservoirs can modify the local water availability and potentially affect the distribution of water bodies, leading to changes in the MNDWI values and influencing the overall vegetation dynamics in the area. Therefore, the combination of climatic factors and human interventions such as the Transboundary Project contributed to the observed variations in the vegetation indices. Sousa et al. [63], in their study of the Terra Nova River Basin, found that there was a direct influence of the release of water from the Terra Nova reservoir of the Transboundary Project on an increase in vegetation cover and expansion of crops along the riverside region. Persistent water features along non-perennial rivers are important resources, providing habitats for plants, animals, and humans during dry periods. These features, which include pools, springs, waterholes, and wetlands, can be sustained by aquifers or runoff [64]. The presence of persistent water along rivers that are typically seasonal is common in different climatic zones around the world, and in Brazil, it is especially used in semiarid regions for artificial perpetuation of rivers through construction of water reservoirs. These human works alter the natural dynamics of water systems, generating anthropic hydrographic basins [65,66].

One of the limitations of our study is that is that this analysis was based on Landsat-8 satellite images, which have spatial resolutions of 30 m. Although these are high-resolution images, using very-high-resolution imagery could provide more detailed information on land cover and land use changes. It is worth noting that very-high-resolution Earth observation data are currently not available in an open-source format for all users [27]. Despite this, important results are being obtained from high-resolution images in the Brazilian semiarid region [4,8,9,35,63,67]. For instance, a study by Sousa et al. [67] on the spatial distribution of the annual precipitation in a Brazilian semiarid basin and its impact on land use and land cover dynamics (vegetation–water nexus) used images with resolutions of 30 m, thus providing information on rainfall patterns and their correlation with vegetation and water availability.

Regarding future research and studies on land cover and land use changes, vegetation dynamics, and water bodies in the semiarid region, it would be interesting to apply this methodology to all areas affected by the São Francisco River Transboundary Project and assess the positive and negative impacts of this endeavor on the region. Another alternative is cloud processing, such as developing scripts in Google Earth Engine (GEE) to analyze the spatial and temporal variabilities of vegetation vigor, canopy structure, and water body dynamics. This can help identify landscapes where intensification of anthropogenic processes is occurring.

Overall, the results of this study have provided valuable information about land cover and use changes, vegetation activity, and water bodies in the study area, which can be used to support decision-making processes related to natural resource management and conservation.

5. Conclusions

In this study, we aimed to analyze the land cover and land use changes, vegetation dynamics, and water bodies in a specific region located between the municipalities of Terra Nova and Cabrobó in the state of Pernambuco, Brazil. Our findings have important implications for understanding the impacts of agricultural activities and the presence of reservoirs, particularly those associated with the São Francisco River Transboundary Project, on the region’s environment. Additionally, we examined the spatiotemporal variations in such vegetation indices as the Normalized Difference Vegetation Index (NDVI), the Leaf Area Index (LAI), and the Modified Normalized Difference Water Index (MNDWI) to gain insights into vegetation vigor, canopy structure, and water body dynamics.

Based on our analysis, we observed significant changes in land cover and land use in the study area over the analyzed period. The expansion of agricultural activities and the construction of reservoirs associated with the São Francisco River Transboundary Project have led to changes in vegetation patterns and the dynamics of water bodies. These changes have important implications for the environment and the sustainability of the natural resources in the region.

The results of our study highlight the importance of utilizing remote sensing techniques and geoprocessing of orbital images for monitoring and assessing of environmental changes in semiarid regions. The use of biophysical indices, such as the NDVI, LAI, and MNDWI, provides valuable insights into the intensification of anthropogenic processes and can guide the development of policies and management strategies to mitigate environmental degradation.

In conclusion, this study contributes to the understanding of land cover and land use changes, vegetation dynamics, and water body dynamics in the semiarid region of Pernambuco, Brazil. These findings have implications for sustainable natural resource management and conservation in the study area. Further research is needed to expand this analysis to other regions and to incorporate very-high-resolution imagery for more-detailed assessments.