Identification of desertified and preserved areas in a conservation unit in the state of Paraíba – Brazil

Cuadernos de Investigación Geográfica Geographical Research Letters 2022 Nº 48 pp. 59-78 EISSN 1697-9540 Copyright © 2021, The authors. This work is licensed under a Creative Commons Attribution 4.0 International License. http://doi.org/10.18172/cig.5098 Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 59 IDENTIFICATION OF DESERTIFIED AND PRESERVED AREAS IN A CONSERVATION UNIT IN THE STATE OF PARAÍBA - BRAZIL LEANDRO F. DA SILVA¹ , BARTOLOMEU I. SOUZA¹ , RAFAEL CÁMARA ARTIGAS²* ¹Dept. Geography, University of Paraíba, Brazil. ²Dept. Physical Geography and Regional Geographic Analysis, University of Seville, Spain. ABSTRACT. The objective of this study is to identify and analyse the main characteristics of areas potentially degraded by desertification and of preserved areas using the Soil Surface Moisture Index (SSMI), alongside the Land Surface Temperature (LST) and Normalized Difference Vegetation Index (NDVI). The study is based on a set of points obtained in the field and from the RGB false colour image for the Environmental Protection Areas (EPA) of the Cariri, in the semi-arid region of Paraíba, using a space-time cross-section covering both rainy and dry periods. The results showed that at all points in Desertified Areas, the main characteristics were a low SSMI, high LST and low NDVI in both periods. The Preserved Areas, on the other hand, presented a high SSMI, moderate LST and high NDVI in the rainy period, with the same characteristics repeated in the dry period for SSMI and NDVI, but with a low LST. Timely identification of these characteristics, both in areas degraded by desertification and in better preserved areas, can provide useful information for future decisions relating to the physical and territorial management of the Conservation Unit. Identificación de áreas desertificadas y preservadas en una unidad de conservación en el Estado de Paraíba - Brasil RESUMEN. El objetivo de este estudio fue identificar y analizar las principales características de áreas potencialmente degradadas por desertificación y de áreas preservadas. Para ello se utilizaron el Índice de Humedad Superficial del Suelo (IHSS), junto con la Temperatura de la Superficie de la Tierra (TS) y el Índice de Vegetación de Diferencia Normalizada (IVDN). El estudio está basado en un conjunto de puntos obtenidos en el campo y en la composición de la imagen de falso color RGB para el Área de Protección Ambiental (APA) del Carirí, en la región semiárida de Paraíba, utilizando un corte espacio-temporal que abarca la estación lluviosa y seca. Los resultados mostraron que en todos los puntos de las Áreas Desertificadas las principales características fueron el IHSS bajo, TS alto y IVDN bajo en ambos períodos. Las Áreas Preservadas, por su parte, presentaron IHSS alto, TS moderado e IVDN alto en la época de lluvias, con las mismas características repetidas en el periodo seco para IHSS y IVDN, pero con TS bajo. La identificación puntual de estas características, tanto en áreas degradadas por desertificación como en las más conservadas, puede aportar información útil para la toma de decisiones futuras relacionadas con la gestión territorial y física de la Unidad de Conservación. Key words: Desertification, Carirí EPA, SSMI. Palabras clave: Desertificación, APA del Carirí, IHSS. Silva et al. 60 Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 Received: 4 Mayo 2021 Accepted: 6 October 2021 *Corresponding author: Rafael Cámara Artigas. Dept. Physical Geography and Regional Geographic Analysis, University of Seville, Spain. E-mail address: rcamar[email protected] 1. Introduction Brazil has the second highest proportion of dry forests degraded by human activity of all the countries in the Americas (Portillo-Quintero and Sánchez-Azofeifa, 2010). Among Brazil’s different ecosystems, the Caatinga, which is the dominant biome in the Brazilian Northeast, is one of the most threatened and disrupted. The main threats to this ecosystem are the illegal extraction of wood for energy and for producing fences, as well as the expansion of farming, which has created large areas of desertified land (Castelletti et al., 2003) and has even affected the Conservation Units (CU) created by the state. These CUs are defined by the Brazilian Ministry of the Environment (MMA) as territorial units with noteworthy natural characteristics, which serve to ensure the representation of important, ecologically viable samples of the different populations, habitats and ecosystems present in the country and to preserve the existing biological heritage (MMAUC, 2020). These areas allow traditional, sustainable, rational use of natural resources to continue, with special rules and regulations in place for local communities to engage in sustainable economic activity. They are created in law by the federal, state and municipal governments after technical studies are carried out on the areas of land in question and local people consulted where necessary (MMAUC, 2020). CUs in Brazil are divided into two main categories: 1) Fully Protected and 2) Sustainable Use. The latter includes the country’s Environmental Protection Areas or EPAs (MMAUC, 2020), where people can make direct use of natural resources, albeit in a carefully controlled manner. Desertification is defined by the United Nations (UN) as the reduction or destruction of biological potential, starting with alterations to natural vegetation by human intervention in arid, semiarid and dry sub-humid regions. These processes are exacerbated by climatic fluctuations (MMA, 2020). In 1994, the UN oversaw the drafting of a global plan to tackle desertification and the effects of drought by adopting effective measures at all levels, based on cooperation agreements and international collaboration (MTERD, 2020). Nevertheless, many of the countries affected by the issue, including Brazil, have been slow to take action and their measures have had little practical impact. According to data from the National Semi-Arid Institute (INSA, 2015)) in Brazil, the area of the country susceptible to desertification covers 1,340,863km² and includes 1,488 municipalities in nine states in the semi-arid region in the Brazilian Northeast. The main hotspots for desertification in Brazil’s semi-arid region are: Seridó in the states of Rio Grande do Norte and Paraíba; Cariris Velhos in Paraíba; Inhamuns in Ceará; Gilbués in Piauí; Sertão Central in Pernambuco; and Sertão do São Francisco in Bahía. These areas are the product of inadequate or non-existent measures to mitigate the interaction between productive activity and the natural resources available in an ecologically fragile environment (INSA, 2015). However, there is still a lack of knowledge of these degrading processes and their extent, and constant updates of this knowledge are required (INSA, 2015). A significant proportion of the damage to natural resources, especially plant cover, is the result of people in regions at risk of desertification seeking to provide for their basic needs. Other causes Identification of desertified and preserved areas Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 61 include commercial agriculture and excessive demand for raw materials for industrial use (Zhou et al., 2015; Bezerra et al., 2020). The impacts of this overexploitation can be seen in escalating soil erosion, especially sheet erosion, and in processes of salinisation in both irrigated agricultural land and non-irrigated land (MMA, 2020). In desertified areas, vegetation is sparse with many dwarf plants and low levels of diversity (MMA, 2020). Therefore, desertification has a severe social and economic impact and is considered to be one of the most significant environmental issues facing the world at this time (Nascimento, 2015; Bezerra et al., 2020). In the Cariris Velhos region, Paraíba, Brazil, the desertification process is very acute and is related primarily to i) geoecological predisposition or the unstable balance resulting from climatic, edaphic and topographic factors; and ii) different types of direct or indirect human activity, which begin with the removal or degradation of plant cover (Sobrinho, 1982; MMA, 2020). Given these characteristics, a wide range of studies have sought to address the issue in this region specifically (Souza et al., 2009; Souza et al., 2015a; Lemos et al., 2020). As this type of degradation progresses, it is increasingly important to understand the causes in order to develop alternative measures and solutions to mitigate the problem and attempt to slow its advance. With this in mind, current technologies can make a significant contribution to studies of desertification and have already been used in research on the issue from a wide variety of disciplines (Sousa et al., 2012; Vieira et al., 2020). In recent years, the number of methods drawing on computational techniques to identify and situate environmental phenomena has also increased, using tools such as remote sensing and geotechnologies applied specifically to arid and semi-arid environments around the world. These methods are based on bio-geophysical parameters obtained via a series of freely available sensors, such as MODIS (Moderate-resolution Imaging Spectroradiometer) on the Terra and Aqua satellites, and TM (Thematic Mapper) and OLI (Operational Land Imager) on the Landsat satellites, which are mostly used to analyse parameters such as land surface temperature (Sousa et al., 2015b; Santos et al., 2020) and different vegetation indices, including the traditional NDVI - Normalized Difference Vegetation Index and the SAVI - Soil-Adjusted Vegetation Index (Aquino et al., 2012; Silva Filho et al., 2020). Other techniques and methodologies have been used to perform important tasks such as quantifying and situating the influence of moisture on the soil. Some of these allow the surface retention capacity, infiltration, evaporation, influence on vegetation and desertification to be estimated, as in studies conducted by Lopes et al. (2011), Francisco et al. (2017) and Inocêncio et al. (2020), based on the interactions between the different indices and bio-geophysical parameters obtained via remote sensing. Against this backdrop, this study aimed to identify and analyse the main characteristics of areas potentially degraded by desertification and preserved areas in a Conservation Unit (CU) for sustainable use, located in the semi-arid region of Paraíba, Brazil, using the Soil Surface Moisture Index (SSMI) proposed by Lopes et al. (2011), the land surface temperature (LST) and the Normalised Difference Vegetation Index (NDVI). 2. Methodology 2.1. Study area The Cariri EPA is situated between the municipalities of Boa Vista, Cabaceiras and São João do Carirí, between the latitudes 07º20’00” and 7º25’00” S and the longitudes 36º25’00” and 36º15’00” O, as shown in Figure 1. Silva et al. 62 Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 Figure 1. Map showing the location of the Cariri EPA and the areas with the most extensive plant cover in this CU. The Cariri EPA is located in the micro-region of Cariris Velhos in the state of Paraíba, Northeast Brazil. This CU for sustainable use was created by the Secretariat of Environment in the State of Paraíba under State Decree 25083 of 8 June 2004 and covers an area of approximately 156 km². The predominant vegetation in the region is Caatinga, which comprises a variety of physiognomies ranging from tree and shrub formations dominated by xerophilous deciduous species, as well as cacti and bromeliads (Ballén et al., 2016; Souza and Souza, 2016; Lima et al., 2017). Some sectors are characterised by differentiated vegetation, comprising species from humid and sub-humid areas of other Brazilian biomes. These are referred to here as Exceptional Areas and their soils and topography result in locally higher water levels, differentiating them from surrounding areas. This can be seen on varying scales across the interior of Northeast Brazil (Melo, 1988; Mello Neto et al., 1985). The climate in the region is classified by Koppen (1931) as semi-arid (BSh) or dry semi-arid, with low rainfall averaging between 400 mm and 500 mm a year (AESA, 2020a) and an average annual temperature ranging from 25 to 27°C, making this the driest region in Brazil (Souza et al., 2009; Silva et al., 2019a). The main soil types found in the region are Regolith Neosol, Litholic Neosol and Chromic Luvisol, as well as Fluvic Neosol on floodplains (Ballén et al., 2016; Silva et al., 2019a). From a geological perspective, the region is made up of granite rock with large rocky outcrops commonly known as lajedos (Lages et al., 2013); one of the most prominent outcrops in this area is the Bravo Pluton, an ellipsoidal stock formed by granitic orthogneisses with a high metamorphic content (Romano et al., 2018). Its surface is ellipsoidal, measuring almost 12 km long and 5 km wide (Sousa and Xavier, 2017), emerging in the centre-east of Paraíba between the municipalities of Cabaceiras and Boa Vista (Romano et al., 2018). Identification of desertified and preserved areas Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 63 In this CU, the most common landforms have a convex, tabular topography. The highest topographies are linked to the large rocky outcrops in the Bravo Pluton region to the West (W), with small mountain groups to the East (E) such as the Caruá Mountains, where elevations exceed 600m (Fig. 2). Figure 2. Map of Altitude. 2.2. Methodological procedures The images used were produced by the Landsat 8 – OLI satellite in April (rainy period) and October (dry period) 2019 and geometrically corrected at the Q1 level (l1T) to all multispectral bands, including band 10 in the thermal infrared region, the TIR (thermal infrared sensor), orbit 215 point 065, freely available from the United States Geological Survey via USGS Earth Explorer. The multispectral images then underwent radiometric correction and were loaded one by one into the ENVI 5.3 software in MTL format. Using the Radiometric Calibration tool, the correction algorithm was applied simultaneously to all bands. Once they had all been radiometrically corrected, the bands underwent atmospheric correction using the FLAASH Atmospheric Correction tool, which employs one of the most advanced algorithms for this type of correction: Moderate Resolution Atmospheric Transmission or MODTRAN. Using the same tool, a series of parameters were set: the type of sensor was set to OLI, the Tropical category was defined for the atmospheric model, the Rural category was set for the aerosol model, 2-band (K-T) was set for aerosol retrieval and 40km for the initial visibility. The next step was to normalise the corrected atmospheric images, automatically balancing the levels of each pixel by applying a Gaussian function using the Band Math tool in ENVI 5.3. 2.2.1. Land Surface temperature (LST) To obtain the land surface temperature (LST), thermal infrared band 10 (B10) from the TIRS sensor with a spectral range of 10.6 -11.19 μm was used for both months, with three key steps required: I. conversion of the grey levels of the thermal band into spectral radiance, via the following equation (eq. 1): Silva et al. 64 Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 𝐿𝐿𝜆𝜆=�𝐿𝐿max𝜆𝜆 − 𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝜆𝜆 𝑄𝑄cal max − 𝑄𝑄cal min� ∗ (𝑄𝑄𝑐𝑐𝑐𝑐𝑐𝑐 − 𝑄𝑄cal min)+𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝜆𝜆 eq. 1 Where: Lλ= spectral radiance (W/m²M.¹·μm¹); Qcal= quantified pixel value and digital level calibration (DN); Qcalmin = minimum value of pixel levels (DN = 1); Qcalmax= maximum value of pixel levels (DN = 255); Lminλ= minimum spectral radiance (1.238 W/ m² sr¹ μm¹); Lmaxλ= maximum spectral radiance (15.303 W/ m² sr¹ μm¹). II. conversion of radiance values into temperature, represented in degrees Kelvin (K), via the following equation (eq. 2): LST = 𝐾𝐾2 𝑐𝑐𝐿𝐿 (𝐾𝐾1/𝐿𝐿𝜆𝜆 +1) eq. 2 Where: LST= land surface temperature; ln= constant of the equation; Lλ= spectral radiance (result of eq.1); K1= calibration constant 1 (607.76 W/ m² sr¹ μm¹); K2= calibration constant 2 (1260.56 W/ m² sr¹ μm¹). The calibration constants (K1 and K2) used were obtained from the metadata (MTL) file in the information package for the bands provided when downloading images from the USGS website. III. conversion of the ST in degrees Kelvin (K) to degrees Celsius (C), via the following expression (exp.1): 𝑆𝑆𝑆𝑆°𝐶𝐶 = 𝑆𝑆𝑆𝑆𝐾𝐾 - 273.15 exp. 1 Where: ST°C= surface temperature in Celsius; STK= surface temperature in Kelvin (result of equation 2); 273.15= temperature equivalent to 0° degrees Celsius. The following tools were used for these three steps: ArcToolbox – Spatial Analyst Tools – Map Algebra – Raster Calculator in ArcGIS 10.3. 2.2.2. Normalised Difference Vegetation Index (NDVI) The Normalised Difference Vegetation Index (NDVI) was proposed by Rouse et al. (1973) and is obtained via the equation (eq. 3.): 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 = 𝜌𝜌𝑁𝑁𝑁𝑁𝑁𝑁 − 𝜌𝜌𝑁𝑁𝑅𝑅𝑅𝑅 𝜌𝜌𝑁𝑁𝑁𝑁𝑁𝑁 + 𝜌𝜌𝑁𝑁𝑅𝑅𝑅𝑅 eq. 3 Where: ρNIR= near-infrared reflectance; ρRED= red reflectance The values vary from -1, which indicates an absence of vegetation, to 1, which shows maximum vegetation. This process used the red (B4) and near-infrared (B5) bands corrected for two months. ArcGis 10.3 software was then used and equation 3 applied via the following tools: ArcToolbox - Spatial Analyst Tools - Map Algebra - Raster Calculator. Identification of desertified and preserved areas Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 65 2.2.3. Soil Surface Moisture Index (SSMI) The Soil Surface Moisture Index (SSMI) was initially suggested by Zhan et al. (2004) and adjusted by Wang et al. (2010) for semi-arid regions in China. It was subsequently adapted by Lopes et al. (2011) for the semi-arid region in Brazil. The index considers biophysical and geophysical parameters from remote sensing, like the LST and NDVI, and consists of three steps: I. application of the SSMI calculation to the LST values obtained (Zhan et al., 2004; Lopes et al., 2011), via equation 4 (eq. 4): 𝑆𝑆𝑆𝑆𝑆𝑆𝑁𝑁𝐿𝐿𝐿𝐿𝐿𝐿 = 𝐿𝐿𝐿𝐿𝐿𝐿max − 𝐿𝐿𝐿𝐿𝐿𝐿 𝐿𝐿𝐿𝐿𝐿𝐿max − 𝐿𝐿𝐿𝐿𝐿𝐿min eq. 4 Where: SSMILST= moisture index calculated directly from LST; LST= land surface temperature; LSTmin= minimum ST value; LSTmax= maximum ST value. II. application of the SSMI calculation adjusted to NDVI (Wang et al., 2010; Lopes et al., 2011) via equation 5 (eq. 5): 𝑆𝑆𝑆𝑆𝑆𝑆𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 = 1 −� 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑚𝑚𝑚𝑚𝑚𝑚 − 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑚𝑚𝑚𝑚𝑚𝑚 − 𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑁𝑚𝑚𝑚𝑚𝑚𝑚� eq. 5 Where: SSMINDVI= moisture index calculated directly from NDVI; NDVI= normalised difference vegetation index; NDVImin= minimum NDVI value; NDVImax= maximum NDVI value. Constant 1 relates to the reversal of values as the higher the NDVI value, the higher the soil surface moisture level (Wang et al., 2010; Lopes et al., 2011). III. The last calculation is then applied to obtain the final SSMI (Lopes et al., 2011), using the average SSMI generated from the relationship between the SSMILST and the SSMINDVI, via equation 6 (eq. 6): 𝑆𝑆𝑆𝑆𝑆𝑆𝑁𝑁 = 𝐿𝐿𝐿𝐿𝑆𝑆𝑁𝑁𝐿𝐿𝐿𝐿𝐿𝐿 + 𝐿𝐿𝐿𝐿𝑆𝑆𝑁𝑁𝑁𝑁𝑅𝑅𝑁𝑁𝑁𝑁 2 eq. 6 Where: SSMI= soil surface moisture index; SSMILST= moisture index calculated directly from LST; SSMINDVI= moisture index calculated directly from NDVI. The final SSMI displays values ranging from 0 for drier surfaces to 1 for damper surfaces (Zhan et al., 2004; Wang et al., 2010; Lopes et al., 2011). All calculations for this step were carried out using ArcGis 10.3 and the tools ArcToolbox - Spatial Analyst Tools - Map Algebra - Raster Calculator. 2.2.4. Fieldwork and analysis The fieldwork was carried out using a GNSS receiver at the collection points. The landscape at each collection point was also described, with particular reference to the use and occupation of the land over time. Images were recorded using a camera. Fieldwork was carried out in May and June 2018 (rainy period) and November 2019 (dry period), covering a large proportion of the Cariri EPA. Silva et al. 66 Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 Back in the office, the points collected in the field with coordinates from the UTM map used for reference were converted individually to Shapefile format before being separated into two categories for analysis: 1) Desertified areas and 2) Preserved areas. Eleven real points (RP) were obtained for Desertified Areas, which were made up of open shrubby Caatinga at different stages of degradation, and fourteen points were obtained for Preserved Areas, comprising closed shrubby arboreal Caatinga and areas of differentiated vegetation referred to here as Exceptional Areas. The field information relating to the points in the Desertified Areas was classified according to the intensity of desertification following the system proposed by Conti (2007) and organised by Araújo and Lima (2019), as shown in Table 1. The 11 points obtained in the field fell into two desertification categories—slight and moderate—as shown in Table 1. Therefore, the term Desertified Areas was used for all points with these characteristics in this study. Table 1. Intensity of desertification. Classification Characteristics Slight Minor deterioration of plant cover and soil. Moderate Significant deterioration of plant cover and appearance of sand nodules. Evidence of salinisation of soils and formation of gullies and holes. Severe Significant expansion of areas prone to gullies and appearance of sand. Expansion of wind erosion. Increase in ruts and holes in the soil. Very severe Almost complete disappearance of biomass. Sealing and intense salinisation of soils. Source: Conti (2007); Araújo and Lima (2019b). To increase and equalise the number of points to be analysed for each category and ensure greater representation of the areas which were not monitored during fieldwork in particular, the points obtained from the false colour images were added between bands B4, R6 and G5 for each period, according to the characteristics of the real points (Table 2). Table 2. Characteristics of the real points (RP) in the false colour images B4, R6 and G5. Real Points Colour Texture Shape Period Rainy Dry Rainy/Dry Rainy/Dry Desertified Areas White, Light Green and Cyan White, Brown and Cyan Smooth, not very rough Regular and Irregular Preserved Areas Colour Texture Shape Period Rainy Dry Rainy/Dry Rainy/Dry Shrubby arboreal Caatinga Red and Black Dark Green and Black Smooth, not very rough Irregular Exceptional Areas Red Dark Red Smooth, not very rough Rectangular, Thin and Irregular Source: Compiled by the authors using data from the Landsat 8 OLI satellite - bands B4, R6 and G5 – USGS Explorer 2019. Identification of desertified and preserved areas Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 67 The points obtained from the images, based on the characteristics of the real points, were labelled IP, taking into account aspects such as the colour, texture and shape of both categories. In this way, 15 points were obtained for Desertified Areas (IP.11-IP.25) and 11 points for Preserved Areas (IP.40-IP.50), which were added to the real points (RP) to produce a total of 50 points (Table 3) for analysis. Table 3. Final points. Desertified Areas Preserved Areas RP.1 – RP.10 RP.26 – RP.39 IP.11 – IP.25 IP.40 – IP.50 3. Results and discussion 3.1. Land Surface Temperature (LST) and Normalised Difference Vegetation Index (NDVI) The first results obtained were the LST and NDVI values, which were calculated and then described for the entire Cariri EPA during the two periods analysed (rainy and dry), as shown in Figure 3. Figure 3. LST and NDVI values. The LST value calculated for the whole EPA for the month of April (rainy period) varied between 18.5°C and 35.5°C. The NDVI displayed values of -1, relating to water bodies, and 1, indicating the maximum biomass. High levels of water bodies and biomass can be found at this time of year across most of the Carirís Velhos, as February and March are the months of greatest rainfall in the region, with precipitation events extending into July (Medeiros et al., 2015; Silva et al., 2018; Sena et al., 2019). The rains directly influence the vigour of plant growth, as most of the plants in the Caatinga are deciduous and display rapid leaf renewal during the rainy period (Amorim et al., 2009) due to the availability of more water. Silva et al. 74 Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 the ST was higher, ranging from 28.6 - 32.5°C. This is due to the fact that these points are located in more open areas of vegetation, which are influenced by smaller, more widely spaced lajedos. Most of the points analysed displayed an NDVI of 0.711 - 1, with lower values found only at RP.28 and IP.46 and 47 (0.577 – 0.710), which are located in areas with a greater presence of dense arboreal Caatinga (another common vegetation type in the region) on steeper slopes in the Caruá Mountains and between the lajedos of Bravo Pluton. During the rainy period, the vast majority of the points analysed in both Preserved Areas saw LST values below 30°C, while temperatures did not exceed 32°C at the other points in these areas. Echoing the findings of Carvalho and Nakagawa (2000) and Souza et al. (2015b), these areas are characterised by improved conditions for germination for the majority of the Caatinga plant species compared to areas with a higher LST, such as the Desertified Areas, especially during the rainy period, where there is an expansion of biomass and greater retention of soil surface moisture due to the availability of water and the type of vegetation found in these areas. Confirming this observation, during the dry period, the 25 points analysed (RP.26 to IP.50) also had an SSMI of 0.7 - 1.0, indicating that there was a high concentration of soil surface moisture at these locations, as Table 7 shows. Table 7. Relationship between the points and the LST, NDVI and SSMI values for the dry period. Point UTM coord. Category LST NDVI SSMI RP.26 788235/9182755 Shrub.Arbo.Caat. 32.5 - 34.5°C 0.324 - 0.390 0.9 - 1.0 RP.27 788475/9182734 Shrub.Arbo.Caat. 34.6 - 36.5°C 0.324 - 0.390 0.9 - 1.0 RP.28 788363/9182567 Shrub.Arbo.Caat. 32.5 - 34.5°C 0.324 - 0.390 0.7 - 0.8 RP.29 802368/9185462 Exceptional Areas 34.6 - 36.5°C 0.391 - 0.514 0.9 - 1.0 RP.30 801848/9185002 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 RP.31 803593/9185805 Exceptional Areas 32.5 - 34.5°C 0.515 - 0.935 0.9 - 1.0 RP.32 799812/9184732 Exceptional Areas 32.5 - 34.5°C 0.391 - 0.514 0.9 - 1.0 RP.33 802026/9184919 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 RP.34 800425/9184939 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 RP.35 802298/9183836 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 RP.36 801832/9185110 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 RP.37 802283/9185583 Exceptional Areas 32.5 - 34.5°C 0.515 - 0.935 0.9 - 1.0 RP.38 800788/9185034 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 RP.39 800065/9184941 Exceptional Areas 32.5 - 34.5°C 0.515 - 0.935 0.9 - 1.0 IP.40 800760/9184679 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 IP.41 798775/9183112 Exceptional Areas 32.5 - 34.5°C 0.515 - 0.935 0.9 - 1.0 IP.42 798018/9182933 Exceptional Areas 32.5 - 34.5°C 0.515 - 0.935 0.9 - 1.0 IP.43 797945/9183390 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 IP.44 797515/9184745 Exceptional Areas 32.5 - 34.5°C 0.391 - 0.514 0.9 - 1.0 IP.45 798861/9185463 Exceptional Areas 34.6 - 36.5°C 0.515 - 0.935 0.9 - 1.0 IP.46 796185/9184299 Exceptional Areas 34.6 - 36.5°C 0.391 - 0.514 0.7 - 0.8 IP.47 800974/9186686 Exceptional Areas 34.6 - 36.5°C 0.391 - 0.514 0.9 - 1.0 IP.48 787737/9182588 Shrub.Arbo.Caat. 32.5 - 34.5°C 0.391 - 0.514 0.9 - 1.0 IP.49 786774/9182608 Shrub.Arbo.Caat. 32.5 - 34.5°C 0.391 - 0.514 0.9 - 1.0 IP.50 786217/9182160 Shrub.Arbo.Caat. 34.6 - 36.5°C 0.391 - 0.514 0.9 - 1.0 Key: RP = Real points collected in the field; IP = Points collected from the RGB image composition; Shrub. Arbo.Caat. = Shrubby arboreal Caatinga. The LST values ranged from 32.5 to 36.5°C, representing the lowest level calculated for this period. Moreover, the NDVI varied from 0.324 to 0.935, indicating high levels of biomass at most of the points during this period of the year, especially in the Exceptional Areas, as shown in Table 4. Identification of desertified and preserved areas Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 75 In the points relating to areas of shrubby arboreal Caatinga, the NDVI oscillated between 0.324 and 0.390 in areas located on the slopes of the Caruá Mountains, where the dominant vegetation is Caatinga scrubland, and from 0.391 to 0.514 at the highest points in this area, IP.48, 49 and 50, where the phytophysiognomy is still denser and more arboreal. In the Exceptional Areas, some points displayed an NDVI ranging between 0.391 and 0.514, such as RP.29 and 32 and IP.44, 46 and 47, where closed Caatinga scrubland is more dominant. However, the remaining points in this area had biomass values between 0.515 and 0.935, located in strips around the lajedos of Bravo Pluton. These areas of high biomass are clearly visible during the dry period in the region (Figure 6), when many evergreen species can maintain themselves due to higher local moisture levels. In the more distant Bravo Pluton area, the predominant vegetation is open shrubby Caatinga, usually deciduous. Comparing the rainy and dry periods for these points in Preserved Areas, the SSMI and NDVI were high in both and only the LST varied. The lowest temperature values were found in the dry period, while moderate temperatures ranging from 25.5 to 28.0oC were observed in the rainy period at most points, with the exception of more open areas where maximum temperatures of 32.5°C were found. In our understanding, these factors are only possible in these areas due to the presence of certain specific natural characteristics: in the case of shrubby arboreal Caatinga, the relative altitude of the mountains is linked to low local levels of anthropisation, while in the Exceptional Areas, the geological and geomorphological conditions generated by the large rocky outcrops in the Bravo Pluton encourage the concentration of moisture. In the latter case in particular, a kind of micro-environment with unique characteristics develops, contrasting with its surroundings in terms of soil type, moisture retention and nutrient availability, as well as soil and air temperature, shade and other important conditions that make these places conducive to vigorous growth of the species colonising them (Lunguinho et al., 2015; Silva et al., 2019b). It is also relevant to note that conservation of the plant cover is facilitated in these areas by the conditions imposed by the topography and geomorphology, making it difficult for human activities making more direct, impactful use of the land to take place. This is the case of the Caruá Mountains, due to their striking topographic relief, and the Bravo Pluton, due to the presence of large rocky outcrops. 4. Conclusions Combining the points used with the biophysical and geophysical data captured by orbital sensors enabled two different types of environmental conditions to be identified and described and a number of highly relevant characteristics to be observed regarding the two categories analysed in this study. In Desertified Areas, factors such as a higher surface temperature, low levels of biomass and low soil surface moisture are present throughout the year, although they are exacerbated by the dry period. Analysis of the points linked to the Preserved Areas revealed the presence of factors that were more conducive to the balance and maintenance of plant cover, with high levels of biomass and a concentration of surface moisture, as well as surface temperatures sufficient for seed germination in the rainy period and lower temperatures than other areas during the dry period. These results may be combined with other information as part of the management process and search for mitigation measures in degraded areas, especially those areas suffering desertification in semiarid regions. In the case of Conservation Units like the Cariri EPA, the methodology employed in this study may be used to inform land-use management plans and, more broadly, to plan a more rational use of the natural resources available in biomass like that found in the Caatinga and other biomes with a similar climate. Silva et al. 76 Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 Acknowledgements The first author is grateful to CAPES/Brazil for the PhD scholarship awarded via FAPESQ, Public Notice no. 003/2016, Case no. 88887.369009/2019-00, and to the Postgraduate Programme in Geography at the Federal University of Paraíba in Brazil. References AESA (Agência Executiva de Gestão das Águas do Estado da Paraíba), 2020a. Bacia Hidrográfica do Rio Paraíba. Available in: http://www.aesa.pb.gov.br/aesa-website/comite-de-bacias/rio-paraiba/ AESA (Agência Executiva de Gestão das Águas do Estado da Paraíba), 2020b Meteorologia – Chuvas. Available in: http://www.aesa.pb.gov.br/aesa-website/meteorologia-chuvas/ Amorim, I. L., Sampaio, E.V.S.B., Araújo, E. L., 2009. Fenologia de espécies lenhosas da caatinga do Seridó, RN. Revista Árvore 33(3), 491-499. Available in: https://www.scielo.br/pdf/rarv/v33n3/11.pdf Aquino, C.M.S., Almeida, J.A.P., Oliveira, J.G.B. 2012. Estudo da cobertura vegetal/uso da terra nos anos de 1987 e 2007 no núcleo de degradação/desertificação de São Raimundo Nonato – Piauí. Revista RA´EGA o Espaço Geográfico em Análise 25, 252-278. https://doi.org/10.5380/raega.v25i0.28013 Aráujo, S.M.S., Lima, E.R.V., 2019. Semiárido Brasileiro e Desertificação. In Araújo, S.M.S. and Lima, E.R.V. (Coord.), Desertificação no semiárido brasileiro e paraibano abordagens conceituais, metodologias e indicadores, pp. 55-62, Bahia: Paulo Afonso. Available in: http://www.sabeh.org.br/?mbdb_book=desertificacao-no-semiarido-brasileiro-e-paraibano Ballén, L.A.C., Souza, B.I., Lima, E.R.V., 2016. Análise espaço-temporal da cobertura vegetal na área de proteção ambiental do Cariri, Paraíba, Brasil. Boletim Goiano de Geografia 36(3), 55-571. https://doi.org/10.5216/bgg.v36i3.44558 Bezerra, F.G.S., Aguiar, A.P.D., Alvalá, R.C., Giarolla, A., Bezerra, K.R.A., Lima, P.V.P.S., Nascimento, F.R., Arai, E. 2020. Analysis of areas undergoing desertification, using EVI2 multi-temporal data based on MODIS imagery as indicator. Ecological Indicators 117, 01-15. https://doi.org/10.1016/j.ecolind.2020.106579 Carvalho, N., Nakagawa, M., 2000. Sementes: ciência, tecnologia e produção. 4ª ed. Jaboticabal: FUNEP. Castelletti, C.H.M., Santos, A.M.M., Tabarelli M., Silva, J.M.C., 2003. Quanto ainda resta da Caatinga? In: Ecologia e conservação da Caatinga. Editora Universitária da UFPE, pp. 719-734, Recife. Conti, J.B. 2007. Clima e meio ambiente. Atual. São Paulo. Francisco, P.R.M., Chaves, I.B., Chaves, L.H.G., Lima, E.R.V., Silva, B.B. 2017. Umidade antecedente e índice de vegetação da diferença normalizada no mapeamento da Caatinga. Agropecuária Científica no Semiárido 13(2), 82-91. http://revistas.ufcg.edu.br/acsa/index.php/ACSA/index Inocêncio, T.M., Neto, A.R., Souza, A.G.S.S. 2020. Soil moisture obtained through remote sensing to assess drought events. Revista Brasileira de Engenharia Agrícola e Ambiental 24(9), 575-580. http://doi.org/10.1590/1807-1929/agriambi.v24n9p575-580 INSA (INSTITUTO NACIONAL DO SEMIÁRIDO), 2015. Relatórios. Relatório do seminário nacional combate à desertificação, degradação das terras e convivência com a semiaridez para redução da pobreza e um desenvolvimento sustentável. Instituto Nacional do Semiárido. Departamento de Combate à Dertificaçao, FAO, Instituto Interamericano de Cooperaçao para Agricultura. Köppen, W.P. 1931. Grundriss der Klimakunde. Walter de Gruyter. Berlín. Available in: https://www.worldcat.org/title/grundriss-der-klimakunde/oclc/2549942 Lages, G.A. Marinho, M.S., Nascimento, M.A.L., Medeiros, V.C., Dantas, E.L. Fialho, D., 2013. Mar de Bolas do Lajedo do Pai Mateus, Cabaceiras, PB - Campo de matacões graníticos gigantes e registros rupestres de civilização pré-colombiana. In: Winge, M., Schobbenhaus, C., Souza, C.R.G., Fernandes, A.C.S., Berbert-Born,M., Sallun Filho, W., Queiroz, E.T. (Edit.). Sítios Geológicos e Paleontológicos do Brasil. Available in: http://sigep.cprm.gov.br/sitio068/sitio068.pdf Identification of desertified and preserved areas Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 77 Lemos, J.E., Souza, B.I., Diniz, M.T.M., 2020. Sistemas, caos e o processo de desertificação no semiárido brasileiro: complexidade e interações. Ateliê Geográfico 14(01), 136-154. https://doi.org/10.5216/ag.v14i1.57004 Lima, F.S., Almeida, N.V., 2017. Dinâmica espaço-temporal da cobertura vegetal na Área de Proteção Ambiental (APA) do Cariri, Paraíba-PB, Brasil. Revista Brasileira de Geografia Física 10(3), 699-721. https://doi.org/10.5935/1984-2295.20170046 Lopes, H.L., Accioly, L.J.O., Silva, F.H.B.B., Sobral, M.C.M., Araújo Filho, J.C., Candeias, A.L.B., 2011. Espacialização da umidade do solo por meio da temperatura da superfície e índice de vegetação. Revista Brasileira de Engenharia Agrícola e Ambiental 15(09) 973-980. https://doi.org/10.1590/S1415- 43662011000900014 Lunguinho, R.L., Souza, B.I., Queiroz, R.T., Cardoso, E.C.M.A. 2015. Influência dos lajedos na composição florística do seu entorno, no sítio Salambaia – Cabaceiras – PB. Revista Equador 4(3), 230-237. Available in: https://revistas.ufpi.br/index.php/equador/article/view/3643/2119 Medeiros, R.M., Santos, D.C., Francisco, P.R.M., Gomes Filho, M.F., 2015. Análise hidroclimática da região de São João Do Cariri-PB. Revista Educação Agrícola Superior 30(2), 59-65. Melo, M.L. 1988. Áreas de exceção da Paraíba e dos sertões de Pernambuco. SUDENE. Recife, 8 pp. Mello Neto, A.V. de, Lins, R.C., Coutinho, S.F.S., 1985. Áreas de exceção do Nordeste brasileiro considerações conceituais. In: SUDENE, Fundação Joaquim Nabuco - Áreas de exceção do Nordeste (Coord). Pernambuco e Paraíba, pp. 1-11. Recife. Available in: http://observatoriogeograficoamericalatina.org.mx/egal3/ Geografiasocioeconomica/Geografiaagricola/01.pdf MMAUC (Ministério do Meio Ambiente. Unidades de Conservação), 2020. Available in: https://www.gov.br/mma/pt-br MMA (Ministério do Meio Ambiente), 2020. In Programa de Ação Nacional de Combate à Desertificação e Mitigação dos Efeitos da Seca PAN-Brasil. Available in: https://www.gov.br/mma/pt-br MTERD (Ministerio para la Transición Ecológica y Reto Demográfico), 2020. Available in: https://www.miteco.gob.es/es/biodiversidad/legislacion/legislacion-y-convenios/conveniosinternacionales/convencion_desertificacion.aspx Nascimento, F.R., 2015. Os semiáridos e a desertificação no Brasil. REDE-Revista Eletrônica do PRODEMA 9(2), 7–26. Portillo-Quintero, C.A., Sánchez-Azofeifa, G.A., 2010. Extent and conservation of tropical dry forests in the Americas. Biological Conservation 143(1), 144-155. https://doi.org/10.1016/j.biocon.2009.09.020 Rouse, J.W., Haas, R.H., Schell, J.A., Deering, D.W., 1973. Monitoring vegetation systems in the Great Plains with ERTS. NTRS – NASA Technical Reports Server. 1, 309-317. Available in: https://ntrs.nasa.gov/api/citations/19740022614/downloads/19740022614.pdf Romano, M.P.C.G., Rocha, J.G., Santos, M.L.F., Souza, M.C.S., 2018. Unidades paisagísticas no semiárido paraibano: porção Leste do Plúton Bravo, Cabaceiras-PB. In: Plataforma Espaço digital (pp. 1-10). Actas del II Congresso Internacional Da Diversidade Do Semiárido. Paraíba, novembro 8-10 de 2017. Available in: http://www.editorarealize.com.br/index.php/artigo/visualizar/33912 Santos, C.V.B., Carvalho, H.F.S., Silva, M.J., Moura, M.S.B., Galvíncio, J.D., 2020. Uso de sensoriamento remoto na análise da temperatura da superfície em áreas de floresta tropical sazonalmente seca. Revista Brasileira de Geografia Física 13(03), 941-953. https://doi.org/10.26848/rbgf.v13.3.p941-957 Sena, J.P.O., Moraes Neto, J.M., Lucena, D.B., 2019. Variabilidade da precipitação em Sumé e São João do Cariri e suas consequências na agropecuária. Revista Brasileira de Climatologia 25, 278-293. http://doi.org/10.5380/abclima.v25i0.65182 Silva, G.S., Silva, W.S., Silva, A.L., Almeida, N.V., Araújo, L.E., 2018. Análise da precipitação da microrregião do Cariri oriental paraibano. Revista REGNE 4(1), 42-57. Available in: https://periodicos.ufrn.br/revistadoregne/article/view/13938 Silva et al. 78 Cuadernos de Investigación Geográfica, 48 (1), 2022. pp. 59-78 Silva, L.F., Souza, B.I., Bacani, V.M., 2019a. Intensidade da ação antrópica na área de proteção ambiental do Cariri paraibano. Caminhos de Geografia 20(71), 364-383. Silva, L.F., Souza, B.I., Bacani, V.M., 2019b. Desempenho comparativo entre classificadores supervisionados no mapeamento de áreas de Ecótono em região de Caatinga. Caderno de Geografia 29(59), 1083-1105. https://doi.org/10.5752/P.2318-2962.2019v29n59p1083 Silva Filho, R., Vasconcelos, R.S., Galvão, C.O., Rufino, I.A.A., Cunha, E.B.L., 2020. Representação matemática do comportamento intra-anual do NDVI no bioma Caatinga. Ciência Florestal 30(2), 473-488. https://doi.org/10.5902/1980509837279 Sobrinho, J.V., 1982. Processos de desertificação no Nordeste do Brasil: sua gênese e sua contenção. Sudene, 101 pp., Recife. Sousa, F.P., Ferreira, T.O., Mendonça, E.S., Romero, R.E., Oliveira, J.G.B., 2012. Carbon and nitrogen in degraded Brazilian semi-arid soils undergoing desertification. Agriculture, Ecosystems and Environment, 148, 11-21. https://doi.org/10.1016/j.agee.2011.11.009 Souza, B.I., Suertegaray, D.M.A., Lima, E. R.V., 2009. Desertificação e seus efeitos na vegetação e solos do Cariri Paraibano. Mercator 8(16), 217-232. Available in: http://www.mercator.ufc.br/mercator/article/view/250 Souza, B.I., Queiroz, R.T., Cardoso, E.C.M., 2015a. Degradação e riscos à desertificação no alto curso do rio Paraíba – PB/Brasil. Revista da Associação Nacional de Pós-graduação e Pesquisa em Geografia (Anpege) 11(16), 201-222. https://doi.org/10.5418/RA2015.1116.0009 Souza, B.I. Macêdo, L.A., Silva, G.J.F., 2015b. Temperatura dos solos e suas influências na regeneração natural da Caatinga nos Cariris Velhos – PB. Revista RA’EGA 35, 261-287. http://doi.org/10.5380/raega.v35i0.41609 Souza, B.I., Souza, R.S., 2016. Processo de ocupação dos Cariris Velhos – PB e efeitos na cobertura vegetal: contribuição à Biogeografia Cultural do semiárido. Caderno de Geografia 26(2) 229-258. https://doi.org/10.5752/p.2318-2962.2016v26nesp2p229 Souza, N.R.L., Xavier, R.A., 2017. A importância dos “lajedos” na paisagem geomorfológica do Cariri Paraibano. In: Os desafios da Geografia Física na Fronteira do Conhecimento. Actas del XVII Simpósio Brasileiro De Geografia Física Aplicada. Campinas, june 28-30 and 02 july, pp. 6561-6566. Available in: https://www.researchgate.net/publication/321304297_A_importancia_dos_lajedos_na_paisagem_geom orfologica_do_Cariri_Paraibano Travassos, I.S., Souza, B.I., 2014. Desmatamento e Desertificação no Cariri Paraibano. Revista Brasileira de Geografia Física 07(1), 103-116. https://periodicos.ufpe.br/revistas/rbgfe/article/viewFile/232905/26887 Vieira, R.M.S.P., Sestini, M.F., Tomasella, J., Marchezini, V., Pereira, G.R., Barbosa, A.A., Santos, F.C., Rodriguez, D.A., Nascimento, F.R., Santana, M. O., Campello, F.B. C., Ometto, J.P.H., 2020. Characterizing spatio-temporal patterns of social vulnerability to droughts, degradation and desertification in the Brazilian northeast. Environmental and Sustainability Indicators 5, 01-09. https://doi.org/10.1016/j.indic.2019.100016 Wang, H., Li, X., Long, H., Xu, X., Bao, Y., 2010. Monitoring the effects of land use and cover type changes on soil moisture using remote-sensing data: A case study in China’s Yongding River basin. Catena 82, 135- 145. https://doi.org/10.1016/j.catena.2010.05.008 Zhan, Z., Qin, Q., Wang, X., 2004. The application of LST/NDVI index for monitoring land surface moisture in semiarid area. IEEE Transactions on geosciences and Remote Sensing 3, 1551-1554. https://doi.org/10.1109/IGARSS.2004.1370609 Zhou, W., Gang, C., Zhou, F., Li, J., Dong, X., Zhao, C., 2015. Quantitative assessment of the individual contribution of climate and human factors to desertification in northwest China using net primary productivity as an indicator. Ecological Indicators 48, 560-569. https://doi.org/10.1016/j.ecolind.2014.08.043