INTRODUCTION
Between 20 and 30 % of world lands are degraded by salinity (Asfaw et al., 2016), which causes environmental problems, principally in agriculture (Gorji et al., 2017). Many are the factors that cause salinity in soils, among them, the excessive activity of irrigation with extraction of high concentrations of mineral salts in water from the phreatic zone (Ávila et al., 2019). The information of studies accomplished to determine the status of salinity in soils, offers a great amount of information required to plan and implement effective programs for their management (Aldabaa et al., 2015).
Salinity of intervening ground has been detected by using remote sensing (Wang et al., 2020). Salinity is not a process that becomes manifest only upon the ground surface, but in all the profile can cause a limitation in the utilization of optic sensors.
Under those circumstances, it is difficult to monitor the chemical composition of the soil with acceptable precision as well as compiling the information to different spatial scales. A multi-temporary study of salinity requires analyzing a long-time period to be representative of the tendencies and magnitudes of the processes of degradation, as well as for defining basic dynamic features, extension and grade of salinization of soils.
Researches accomplished under different edaphological conditions of use and management of soils have been supported in the use of remote sensing in studies of salinity and its relation with the values obtained in farm and laboratory measurements; in the elaboration of thematic maps for the comparison of periods of time, seasons of the year and in different stages of growth and development of cultivations (Al-Khaier, 2003; Heidinger, 2008; Polishing et al., 2010; Tighten et al., 2016; Hao et al., 2018 ).
The objective of this investigation was to determine the spectral index related with the salinity of the soil with the remote sensing in two epochs of the year contrasting due to their regimens of precipitations, in Mayarí Agroecosystem, at Holguín Province, Cuba.
METHOD
Mayarí Agroecosystem is located at Holguín Province and comprises a surface of 1304,2 km2 (Figure 1) with the biggest agricultural pole of the province and a great part of low irrigation, supplied by Mayarí Dam and the works at the East - West Water Transfer Chanel.
The origin of spectral index utilized for the determination of salinity are images of satellite Landsat 8 OLI (Operational Land Imager) and TIRS (Thermal Infrared Sensor) collection 1, both downloaded from the Geological Service of the United States (USGS) in the grid 011/046. The Survey of Oran topographical sheet was geographic referenced using the Geographic Universal Transverse Mercator (UTM) 18N coordinate system and the WGS84. For its accumulation and the time, Landsat imagery have a space resolution of 30 m, temporary resolution of 16 days and coverage of 185 km of land.
In Table 1, the characteristics of Landsat imagery utilized for the determination by sensor remote of spectral index related with soil salinity in different epochs of the year (May and December) are illustrated. Each image represents a data set grouped in pixels.
Image | Acquisition date |
---|---|
LC08_L1TP_011046_20160519_20170324_01_T1 | May 19, 2016 |
LC08_L1TP_011046_20181219_20181227_01_T1 | December 19, 2018 |
The Software QGis 3.10 was utilized for the determination of spectral index, the realization of thematic maps and radiometric correction, in order to decrease the atmospheric, radiometric and topographic effects in both imagery utilized.
Spectral indices related with the salinity of the soil that were determined are referred in Table 2:
Espectral index | Expression | Reference |
---|---|---|
Salinity Index (SI) |
|
Khan |
Normalized Difference Vegetation Index (NDVI) |
|
Rouse |
Enhanced Normalized Difference Water Index (ENDWI) |
|
Chen |
Normalized Difference. Drought Index (NDDI) |
|
Gu |
Vegetation Soil Salinity Index (VSSI) |
|
Dehni y Lounis (2012) |
B2: Blue band; B3: Green band; B4: Red band; NIR: Near Infrared spectral band; SWIR2: Shortwave Infrared.
Data of Guaro Weather Station of Mayarí Municipality for the years 2016 and 2018 were taken like reference for the study of the epochs with bigger affectation of the contents of salts in the ground and the stress induced by the intense drought. There were 70,6 mm accumulate of average rain in May 2016 and 79,0 for December 2018, which shows a minimum difference between both contrasting epochs.
RESULTS AND DISCUSSION
The descriptive statistics of the spectral values that the pixels take in both epochs of the years examined are shown in the Table 3. The deficit of water shown in the spectral information of December 2018, expressed by the index NDDI with mean values of -0,821, did not have negative incidence in the indices IS, NDVI, ENDWI and VSSI provided that they yielded minor presence of salts in the soil than the spectral information of May 2016 which shows a higher SI (-0,409) with values closer to -1. According to Elhag (2016), the values of the spectral index that indicates the state of salinity of the soil, oscillate from -1 (high presence of salts in the soil) to 1 (reduced presence of salts in the soil).
Epoch of the year | Spectral Index | Minimum Value | Maximum Value | Mean Value | Standard deviation |
---|---|---|---|---|---|
May/2016 | IS | -0.2409 | 0.2187 | 0.0786 | 0.0517 |
ENDWI | -0.1753 | 0.5994 | 0.2858 | 0.1360 | |
NDDI | -269764.2812 | 220090.8281 | 0.1345 | 297.6306 | |
NDVI | -0.3532 | 0.6433 | 0.2661 | 0.1771 | |
VSSI | - 418727.0 | -48510.0 | -74437.2212 | 27351.1005 | |
December/2018 | IS | -0.2893 | 0.5828 | 0.2661 | 0.1493 |
ENDWI | -0.1255 | 0.5555 | 0.2596 | 0.1222 | |
NDDI | -1141697.5 | 163069.7343 | -0.0821 | 1031.6220 | |
NDVI | -0.2893 | 0.5828 | 0.2936 | 0.1493 | |
VSSI | -197326.0 | -44772.0 | -54747.2184 | 5630.8296 |
In December 2018, there were NDVI for zones with vegetation mean values of 0,293 bigger than in May 2016 with 0,2661. The interval of values once the NDVI was gotten, varied from (- 1) to (1) and only the positive values corresponded to zones of vegetation. The negative values generated for a bigger reflectance in visible than in infrared corresponded to clouds, snow, water, zones of naked ground and rocks. The NDVI values can vary in terms of the use of the soil, phenological station, hydric situation of the territory and typical weather of the zone. These properties make that the NDVI had constituted a valuable tool for the evaluation of vegetable covers, as well as for going into the classification and vegetable dynamics and its phenological aspects (Chen et al., 2005).
Martínez et al. (2011) refer that drought, salinity and extreme temperatures are the principal types of stress that cause adverse effects in the growth and productivity of cultivations. The NDDI and the ENDWI, both indices related with the presence and quantity of water in the soil determined from Landsat imagery, allow establishing relations with the presence of salts in the soil according to the study performed by Khan et al. (2005).
The remote sensing of the salinity as NDVI has been largely utilized, granted that it indicates the status of the vegetation according to the stress for salinity (Kumar and Aldakheel, 2014). Figure 2 specifies the maps of SI and NDVI in May 2016 and December of 2018 in the agroecosystem of Mayarí.
The visualization of contrasts in the tonalities that take the salinity and the vegetation in this area gives the appearance of the variation of these measures in both epochs.
Platonov et al. (2013) refer that for identifying the status of the vegetation in front of the saline stress it becomes necessary the analysis of large extensions due to the variation of land use and management by the producers (Jacobus, 2017).
The presence of water in the soil has a great influence on the state of salinity, therefore, the determination of the indices related to the stress caused by drought in the plants suggest that there was variation in the months of May 2016 and December 2018 in Mayarí agroecosystem (Figure 3).
The obtained results agree with that presented by Oliva et al. (2008), who refer that drought is major at the dry and hot regions, and existing a bigger concentration of salts in the upper layer of the ground due to the evapotranspiration, that exceeds precipitation.
CONCLUSIONS
The results obtained in the investigation highlight the precision of the remote sensing in the determination of soil salinity and of index spectral related with the stress in the plants caused by this process of degradation. The imagery utilized for the analysis of spectral index showed a bigger salinity index in the month of May 2016, what could be given by the starting of the rainy season and the month of December 2018, the ending of the rainy season at the country. The maps manufactured from the spectral information show the contrast in both years examined from their index of salinity, of vegetation and their relation with the stress for intense drought.