Soil organic matter (SOM) is an important term to realize soil productivity and quality that is extremely influential on soil physical, chemical and biological processes; SOM is one of the key soil properties controlling nutrient budgets in agricultural production systems and is an important index of soil productivity. Remote sensing (RS) and Geographic Information System (GIS) techniques were used to assess organic matter in soil and determine the relationship between measures SOM in field and digital data to calculate or obtain the correlation coefficients applied to evaluate the strength and direction of the linear relationships. In this study Normalized Difference Vegetation Index (NDVI), Soil Adjusted Vegetation Index (SAVI) and Bare Soil Index (BSI) were used. The results show that the relationship between vegetation indices (NDVI, SAVI) and SOM in whole study area was (R 2 = 0.19, p < 0.05), while the relationship in arid areas was (R 2 = 0.01, p < 0.05), and the relationship in semi-arid areas was (R 2 = 0.13, p < 0.05). It can be concluded that the diversity in vegetation cover and humidity effects to the relationship between vegetation indices (NDVI, SAVI) and SOM, where these relationships increase rapidly in semi-arid areas more than arid areas. In the other hand about the relationship between SOM and BSI (R 2 = 0.11, p < 0.05), soil organic carbon increases with increasing NDVI and decreasing BSI. NDVI, SAVI and BSI were considered a useful index to detect the spatial distribution of SOM concentrations and mapping using remote sensing data.
Soil organic matter (SOM) has a significant effect on the soil processes which in turn affects the ability and productivity of the soil. Moreover, it has an impact on water capacity of soil, soil structure and nutrients which increases soil quality [
Soil organic matter is a mix component which leads to change soil characteristics and affected by land use, climate, vegetation and soil type. That means the productive ability of agriculture will be compromised by deterioration in soil physical properties if SOM concentration decreased in soil [
The quantity of soil organic matter determines the input of organic matter and its rate of oxidation, soil texture, climate, mineralization rate and organic matter decomposition. The clay helps to settle down and limit SOM decomposition. Therefore, soil texture is considered an important factor to agricultural crops as well as rainfall and temperature is a significant factor to the agricultural crops because of its high impact on the OM decomposition rate [
SOM includes all OM in the soil except in decayed material. Thus, SOM supports plant because OM acts as an important source of soil nutrients to plants. The OM is reduced by agricultural and economic development [
The prime interest for soil experts and environmental directors is studying soil characteristics effectively and opportunely. With the growing request of locative and temporal decision in soil characteristics in different agriculture sectors, traditional laboratory techniques were proved to be unsuitable [
Moreover, traditional methods are not always available because of lacking both time and cost connected with sample aggregation. So, it was focused on the employ of remotely sensed data to quantify variation in soil physical characteristic [
Soil scientists are interested in using remote sensing and its applications in soil surveys and soil mapping especially after the evolution of optical sensor in conjunction with field measurements [
This study aims to use simple method of results from satellite data and determine the relationship between measured SOM in field and obtained data using remote sensing techniques, find correlation coefficients which applied to evaluate the strength and direction of the linear relationships between variable couples [
Study area is located northeast and northwest Jordan and distributed in two governorates: Mafraq and Irbid governorates
The average monthly overflow volumes are in December, January, February
and March. Evaporation rate constitutes about 90% of the aggregate rainfall [
The study area consists fundamentally of Paleogene to quaternary continental tuffs and basalts, outcrops back to the Azraq Formation, existing within a rift zone, limited by the NW–SE trending Fuluq and Sirhan faults [
The methods implemented in this study can be broadly divided into three subsections; fieldwork, image processing and statistics analysis.
Soil Samples from 22 sites located in two zones (semi-arid and arid zones) were carefully collected in plastic bags and appropriately labeled. The soil samples were collected from depth 30 cm. The fieldwork visit includes three types of lands; available agricultural in the study period, abandoned land, and protected land. The purpose of this fieldwork was to collect soil samples from these locations in validation of the land degradation index by assessing soil quality and structure in addition to chemical analyses of these samples. Samples were then classified to different soil texture type by sieve analysis method [
C % = [ ( B − S ) × M of Fe 2 + × 12 × 100 ] weightofsoil ( g ) × 4 (1)
B is the volume of Fe2+ solution used to titrate blank (mL). S is the volume of Fe2+ solution used to titrate sample (mL), and 12/4000 is milliequivalent weight of C in grams to convert easily oxidizable organic C to total C, divided by 0.77 (or multiply by 1.30) or other experimentally determined correction factor. Therefore, the percentage of organic matter can be estimated by the following Equation (2):
OM % = C % 0.58 = C % × 1.72 (2)
SOM was classified for five classes and used to implement this study after a modified soil-organic-matter test [
The satellite used Landsat 8 OLI (Operational Land Imager) and TIRS (Thermal Infrared Sensor) used in this study in May 2017 path 169 and row 45. The image was obtained from the United States Geological Survey (USGS) Global Visualization (GloVis) site and geometrically corrected and rectified to UTM zone 36. In order to prepare data, managing and analyzing software were conducted as follows: ERDAS IMAGINE 2014 in this study ERDAS was applied in importing and enhancement, as well as ArcGIS 10.5 used to digitizing, indexing and image analysis, geo-referencing, creation of database. Arc map was used for the composition and generation of maps.
Three indices were applied in this research to achieve the aims, where these indices are widely used and these were selected to suit the acquisition data that used in this paper. The Normalized Difference Vegetation Index (NDVI)—which is a normalized ratio of red and near-infrared reflectance. Rouse et al., 1973 have been used in many phonological studies. NDVI is calculated as Equation (3) [
NDVI = λ NIR − λ R λ NIR + λ R (3)
where the (λNIR) represent reflection in near infrared, IR (λ ≈ 0.8 µm) regions of the spectrum. (λR) represent surface reflectance in the visible light, R (λ ≈ 0.6 µm, “red”). The large variations in NDVI can be due to variations in soil brightness [
- | Class | Description | Ranking |
---|---|---|---|
Organic matter (%) | a | <1 (mineral soil, very low organic-matter content) | Very low |
b | 1 - 3 (mineral soil, moderate organic-matter content) | Low | |
c | 3.1 - 15 (mineral soil, high organic-matter content) | Moderate | |
d | 15 - 30 (organic soil) | High | |
e | >30 (organic carbon soil) | Very high |
reason of considerable differences in NDVI values and that it’s not represent of the actual vegetation cover [
The NDVI is measured by satellite sensors. Rouse et al. (1973) [
The Soil Adjusted Vegetation Index (SAVI) (Huete, 1988) [
SAVI = λ NIR − λ R λ NIR + λ R × ( 1 + L ) (4)
where the NIR represents reflection in near infrared (730 - 1000 nm), RED represents surface reflectance in the visible light (550 - 700 nm) and L is soil adjustment factor (Ray et al., 2001). The value of L varies by the amount or cover of green vegetation and varying from 0 (close canopy cover) to 1 (open canopy cover): in very high vegetation regions L = 0; and in areas with no green vegetation, L = 1. Generally, an L = 0.5 works well in most situations and is the default value used.
However, removal of plant covers jeopardizes the carbon cycle and decreases the percentage of organic matter in the soil, as well as resulting in a degradation of physical soil properties and structure [
Bare Soil Index (BSI) value was estimated using combines blue, red, green and near-infrared bands to detect the soil changes of the Landsat 8 OLI and TM data to determine the condition of the uncovered soil by vegetation which is helpful to determine the state of the soil organic matter on the region. BSI is described by the following Equation (5) according to Jamalabad & Abkar (2004) [
BSI = [ ( λ R + λ G ) − ( λ R + λ B ) ( λ NIR + λ G ) + ( λ R + λ B ) × 100 ] + 100 (5)
Parameter | Class | Description |
---|---|---|
Vegetation Removal (%) | 1 | Low, <1.5 |
2 | Moderate, 1.5 - 2.5 | |
3 | High, >2.5 |
Results of laboratory measurements of soil indicate that organic matter percentage in the soil (Organic Carbon) in study area varies from low to very low based on
It was found that the relationship between Vegetation Index (NDVI, SAVI) and SOM was statistically significant and the (R2 = 0.19, p < 0.05), as shown in
SOM = 2.96 × NDVI + 0.36 (6)
This model describes the relation between (NDVI, SAVI) and SOM in both arid and semi-arid regions. However, the relationship is relatively different in arid areas from that in semi-arid areas (R2 = 0.01, p < 0.05), (R2 = 0.13, p < 0.05), respectively. It was also found that the relationship between (NDVI, SAVI) and SOM increases in semi-arid areas that have diversity in vegetation cover and humidity more than arid areas that have dry climate change but do not have variety in vegetation cover. The means of laboratory measurement of soil in both arid and semi-arid or humidity regions were (0.68, 0.89), respectively.
The relationship between vegetation index such as NDV and BSI is represented
Parameter | SOM | NDVI | SAVI | BSI |
---|---|---|---|---|
MIN | 0.38 | 0.08 | 0.12 | 100.11 |
MAX | 1.68 | 0.19 | 0.28 | 102.96 |
MEAN | 0.78 | 0.14 | 0.21 | 101.32 |
SD | 0.27 | 0.04 | 0.06 | 1.01 |
S.DE | 0.06 | 0.01 | 0.01 | 0.24 |
BSI | SAVI | NDVI | SOM | |
---|---|---|---|---|
1 | SOM | |||
1 | 0.1874 | NDVI | ||
1 | 1 | 0.1874 | SAVI | |
1 | 0.8114 | 0.8114 | 0.1136 | BSI |
in
It was found that there are differences in the relationship between measured SOM and Bare soil Index (BSI). This relation is shown in
SOM = 0.09 × BSI − 9.99 (7)
The outcomes also indicated that SOC increases with increasing NDVI and decreasing BSI. Finally, the multiple regression models were fitted into ArcGIS to generate the digital SOC map using raster calculator
It was also found that the regions, that have both a high rainfall (250 - 400 mm annually) and a good vegetation percentage, have the high concentrations relatively of SOM. These regions were distributed in eastern part boarders and some regions in the central part of study area which is semi-arid. The region covered by urban, rock and sand has extremely low SOC values about zero, while the central part and southern-east boarders have low of SOC values and have a low rainfall relatively (less than 200 mm annually) and a high temperature.
The SOM measured values showed a good correlation with used indices, where
this study suggested that these vegetation indices (NDVI and SAVI) and bare soil index (BSI) are helpful to detect the spatial distribution of SOM concentrations. In addition the statistical analysis results were useful to analyse measured values in the field and to understand the relationships that helped to estimated SOM concentrations by remote sensing data, although the pixel size (30 m × 30 m) used in the study area was relatively large with this area, which reflected the variation in SOM concentrations. The results of this study support the approach that the remote sensing data are useful to detect and investigate the environmental changes. This approach was used through mapping and determines the spatial distribution of SOM concentrations using spectral indices processed remote sensing data; techniques could be effective to monitor and manage soil.
Authors are thankful to anonymous reviewers for their constructive comments and suggestions to improve the manuscript. Authors also are thankful to Department of Earth and Environmental Sciences members in Al Al-Bayt University for providing all necessary support.
The authors declare no conflicts of interest regarding the publication of this paper.
Ibrahim, M., Ghanem, F., Al-Salameen, A. and Al-Fawwaz, A. (2019) The Estimation of Soil Organic Matter Variation in Arid and Semi-Arid Lands Using Remote Sensing Data. International Journal of Geosciences, 10, 576-588. https://doi.org/10.4236/ijg.2019.105033