CC BY
28ECOSYSTEMS: ECOLOGY AND DYNAMICS, 2021, Vol. 5, No. 3, P. 34-59
= STRUCTURAL ORGANIZATION OF ECOSYSTEMS AND PATTERNS =
OF THEIR DISTRIBUTION
UDC 631.4
AGRICULTURAL LANDS IDENTIFICATION ON THE SATELLITE IMAGERY1
© 2021. I.N. Gorokhova*, I.N. Chursin**, N.B. Khitrov*, Ye.I. Pankova*
*V.V. Dokuchaev Soil Science Institute Russia, 119017, Moscow, Pyzhevsky Per. 7, Building 2. E-mail: [email protected]
**Geoinformation Center of the Russian Academy of Sciences Russia, 119019, Moscow, b/o 168, Novy Arbat 11. E-mail: [email protected]
Received Jili 30, 2021. After revision August 30, 2021. Accepted September 01, 2021.
Our work aims to determine the possibilities of identification of the fields with various crops, leys and fallows, using the multispectral high resolution (10 m) satellite imageries taken by Sentinel-2 Satellite. For this we studied the irrigation system of Volga-Don in the dry steppe zone of Volgograd Region. We analyzed the images, taken from February to August 2020, and used the field materials, obtained from the surveys that took place from August to September 2020 in the said area. The variations of the agricultural lands structure were carried out visually and automatically, which allowed us to identify winter cereals, spring crops, leys and fallows. Among them we identified such crops as Sudan grass, corn and soybeans. The accuracy of automated method of structure identification was 75%. We used a combination of different channels for multispectral images to divide the irrigated/non-irrigated fields of the irrigation system and the fallows of different years.
This resulted in a creation of a map of agricultural lands structure for the central part of the Volga-Don irrigation system. It shows the cultivated crops on the irrigated and dry lands, leys, uneven-aged fallows and the exact area of each field.
The calculated NDVI values represented the state of crops during different stages of their growing season. The obtained data made it possible to select key sites and study the irrigated soils, because the agricultural crops condition reflected the structure and degradation of the land cover and helps to select the required sites properly.
Our studies have shown us the possibility to use the high resolution remote data for assessment of the structure and state of agricultural lands, and selection of the key sites for soil surveys, the information about which is necessary to develop and select the suitable ameliorative measures. Keywords: structure of agricultural lands, crops, leys, fallows, irrigated lands, satellite imagery processing.
DOI: 10.24412/2542-2006-2021-3-34-59
In the XX century the large irrigation systems were constructed in the dry steppe zone of European Russia, which made it possible to secure the food supplies for the country even during the dry years. The most active construction took place in the 1950s-1960s; by the middle of the 1980s, the developed areas of the irrigated lands reached their maximum.
However in the 1990s, the large irrigation rates, lack of drainage and water proofing in the most irrigation channels caused a rise of groundwater level, a development of secondary salinization hotbeds, soils alkalization, flooding, irrigational erosion and other negative consequences. The following economic crisis reduced the area of irrigated lands and changed the management of
1 The study was carried out according to the Governmental Tasks No. 0591-2019-0023 "To Study the Influence of Natural Conditions and Land Use Features on Soil Cover, Including Participation of Saline Solonetzes and Slitized Soils, and to Develop an Assessment Technology for the Intense Soil Exploitation in the Agricultural Lands". and No. AAAA-A19-119012390065-9 "Development of Methods for Analysis and Complex Use of Earth Remote Sensing Data, Based on the Modern Geoinformation Technologies, for Monitoring of the Natural Landscapes in the Agricultural Lands, Using Groups of Small Spacecraft".
hydraulic systems. Large areas were abandoned and eventually turned into perennial fallows, some areas were used for dryland farming, and some were irrigated again only after a long period of abandonment or dryland farming, while the other changed their irrigation method. Such drastic transformations of the irrigated lands had affected the state of their soils.
These days the government aims to reconstruct the irrigation systems to restore and improve the management of the ameliorative complex, as well as the amelioration of irrigated lands. To achieve this, it is very important to understand the tendency of soils and soil cover of irrigation systems to transform and evolve over a long period of functioning, and take into account their lithological and geomorphological features, duration of irrigation period and intervals in it. The knowledge, gained from the study of the present state of irrigated soils, will allow us to purposefully select and carry out the necessary agricultural and ameliorative measures to restore the fertility of the soils.
The soil cover of the dry steppe, solonetz zone of the Volgograd Region is heterogeneous, therefore, the surface of agricultural fields is represented on the satellite images by combination of spots, variously sized and colored. These spots are genetically related to the soil structure and affect the crops differently, leading to their sparseness or loss in some cases. To understand the effect of soils and associated spatial heterogeneity of fields on a particular crop, as well as to determine how much the crop condition reflects the condition of soil cover, it is necessary to know the structure of agricultural lands. This allows us to select the key sites for a survey of irrigated soils and take adequate ameliorative measures. The high resolution satellite imagery, being an open and periodic source of information, makes it possible to carry out fast studies in the large areas of irrigated soils.
The aim of this research is to identify the possibilities of recognition of the agricultural lands (fields with different crops, leys and fallow) structure on the high resolution satellite images (10 m), using the visual and automated methods, while taking into account the state of crops and fields, in order to calculate the areas and select the key sites for a further soil survey.
Until 2020 we were studying the agricultural land and irrigated soils in the Volgograd Region, in the area of the Svetloyarsk irrigation system with cultivated fodder grasses, winter cereals and vegetables. I.N. Gorokhova and E.I. Pankova (2017), I. N. Gorokhova et al. (2020) wrote about usage of satellite data to recognize those crops. It was stated that Landsat-8 images (maximum resolution - 15 m) could be used for May or June, because fodder grasses and ripened winter cereals were well-defined on them. However, the Volga-Don system has a wider range of crops, which requires a different approach to the analysis of remote sensing data. For this a series of satellite images taken in different months was used.
The dry steppe zone of Russia undergoes an active agricultural land use. These soils were studied by great Soviet and Russian scientists and by the staff of the V.V. Dokuchaev Soil Science Institute in particular. They recorded the natural state of soils and soil cover in their original condition, before the territory was active developed and irrigated (Ivanova, 1928; Kovda, 1937; Rode, 1947; Antipov-Karatayev, 1953; Bolshakov, 1961).
The irrigated soils have been studied for a long time, in many fields, which results in a creation of many works. These studies are aimed to explore the fluctuations and rise of the groundwater level during irrigation (Minashina, 1978; Zaidelman, 1993), irrigation erosion of irrigated soils (Kuznetsov et al., 1990; Kozlovsky, 1991), secondary salinization processes (Kovda, 1946; Zimovets, 1981; Shahid et al., 2012; Zaman et al., 2018; Ren et al., 2019; Bahmaei et al., 2020), degradation of soil structure at the macro and micro levels (Khitrov et al., 1994; Prikhodko, 1996), alkalization processes (Zimovets, 1981; Zinchenko et al., 2020); carbonization processes (Baranovskaya, Azovtsev, 1981; Lyubimova, Degtyareva, 2000; Lyubimova, Novikova, 2016; Sizemskaya, 2013; Gorokhova, Pankova, 2017), irrigation water effect on soil properties (Bezdnina, 1997; Dedova, 2018; Zinchenko et al., 2020), and movement of soil moisture in irrigated soils (Jiménez-Aguirre et al., 2018).
Aside from a traditional ground sampling, many other methods for assessing the state of irrigated
soils are used. Since the 1950s of the XX century, they have been primarily based on the satellite data (Simakova, 1959; Myers et al., 1966; Pankova, Mazikov, 1975; Vyshivkin, 1976; Antonova, Kravtsova, 1976; Richardson, 1976; Sinanuwong et al., 1980; Andronnikov, 1979; Kharitonov, 1982; Manchanda et al., 1983; Myers, 1983; Mamedov, 1985; Pankova, Soloviev, 1993; Hick, Russell, 1990; Singh, 1994; Dwivedi, 1996), and still use it nowadays (Rukhovich, 2009; Konyushkova, 2014; Iqbal, 2011; Abbas et al., 2013; Savin et al., 2014; Pankova et al., 2014, Gorokhova, Pankova, 2017). The majority of articles are written about the saline and solonetz soils.
A large number of the newest literary sources describe the irrigation and its consequences worldwide, while being based on satellite data. The studies of soil salinity were carried out in the Middle East (Allbed et al., 2014), North Africa (Hihi et al., 2019), Turkey (Gorji et al., 2017), China (Jiang, Shu, 2018; Wang et al., 2019; Chi et al., 2019; Ren et al., 2019) and Russia (Komissarov et al., 2019; Gorokhova et al., 2018-2020). The data of Landsat ETM +, Landsat-8 OLI and Sentinel-2 is the most widely used (Jiang et al., 2018; Masoud et al., 2019; Hasanlou, Eftekhari, 2019; Gorokhova et al., 2020).
A group of scientists (Hassania et al., 2020) compiled maps to forecast the distribution of saline and alkaline soils worldwide, using matrix and regression models of the spatiotemporal variability of soil salinity and alkalinity (1980-2018). Similar study was carried out for Hungary (Szatmari et al., 2020).
During the visual decryption of the images, the saline soils can be recognized by white spots, created by a salt crust on their surface. However, this information does not distinguish soil salinity throughout the entire root-inhabited layer, thus limiting the interpretation possibilities. Therefore, during the analysis of multispectral images, salinity is often associated with the state of the crops (Gorokhova et al., 2019). To distinguish different crops, irrigated and non-irrigated fields, to solonetzic, saline and non-saline soils the combinations of various channels of satellite imagery are used, as well as calculated and experiential indices (Allbed et al., 2014; Hihi et al., 2019).
The most used and identified index in the satellite imagery is Normalized Difference Vegetation Index (NDVI), which was first described in 1973 (Rouse et al., 1973). It is a simple index for a qualitative and quantitative assessment of the green mass of vegetation cover. NDVI can be calculated as NDVI=(NIR-RED)/(NIR+RED), where NIR is a brightness or reflectance in the nearest infrared area of the spectrum (0.7-1.0 p,m), and RED is in the red area of the spectrum (0.60.7 p,m). It presumes the ratio of the maximum plants absorption of solar radiation from the red area, and the maximum reflection of the leaf cellular structures in the infrared one.
There are some other vegetation indices, the most well-known of which are SAVI (Huete, 1988), ASVI (Crippen, 1990) and ARVI (Kaufman et al., 1992). They use different channels or corrective coefficients.
However, some researchers do not trust such indexes. A number of authors (Koroleva et al., 2017) believe it is incorrect to use NDVI in determining of the area of the open soil surface, since its accuracy is only 65%. In their opinion, the reliable identification of the area in the spectral space is possible, if the original technology for the spectral vicinity of the soil line, which was developed by them, is used, because its accuracy is 90%. According to this method, the flat part of the tasseled cup, which is considered a soil line (Kauth and Thomas, 1976), is not a synonym with an area of open soil surface, and occupies a significantly smaller area than the area of a soil line.
Nevertheless, vegetation indices are widely used in data services to vegetation monitoring. One of them is VEGA-PRO (2020), which is a satellite service that uses time series of vegetation indices and allows to analyze condition of the vegetation cover (crops and forests), along with its seasonal and long-term dynamics for any point or region of Russia. This system was created by the Russian Space Research Institute of the Russian Academy of Sciences.
Generally speaking, the review of literary sources indicates that there is no strictly defined combination of channels or indices for determining the state of vegetation, crops and soil salinity for different territories. For each specific case an individually selected approach should be developed.
Materials and Methods
Volga-Don irrigation system is located in the dry steppe zone of the Volgograd Region, in the south of the Volga Upland, stretching along the Volga-Don channel, west of Volgograd. We studied its central area, occupied with cultivated fields and fallows. The total area of the study is 8988.9 ha (Fig. 1).
* ■ * -A ¡10^ Sr Uk | 1 L * ' »■ . * * 1 H tlw*l , » 4 1
-1 tk ■! r*. -» ■ *i /jt, ^ - _
VOLGOG R A D J
<1 if i'
Volga-Don irrigation system % f iv ' 5 IfT . . ' « T , N V ¡¡^r /
|Lk 3—«;■ ./ ' 4I C{, . »
f ES- -toi ' v ■ sEyVi? ■ >lJl ' ■ V lik^ ^ ^^ - -V A N
Novy Rogachik 'y A c aH VOLGOGRAD
Volga-Don irrigation system
A ™ Gorny
Varvarovskoye 5oogic F.irtb reservoir A, « 'N Kft _i-'_1
Fig. 1. Volga-Don irrigation system on a satellite image from Google.Earth, 5/11/2020.
The south area, where the irrigation system is located in the Volga Upland, was entirely covered by the Yergeni deposits during the Pliocene Epoch, and then it underwent a gradual denudation. Currently its watersheds are formed with the Yergeni sands, and red-brown Scythian clays and loess loams to the south. The soil-forming rocks are yellow-brown, and their sandy loams are of various depth. The groundwaters are deep and usually slightly mineralized, with a hydrocarbonate-sodium composition. The soil cover is represented by chestnut solonetzic complexes, with dominating loamy chestnut alkaline soils and a small addition of solonetz
(Degtyareva, Zhulidova, 1970; Zinchenko et al., 2020). The variously salinity of alkaline soils and solonetzes affects the state of developing crops.
In this study we used multi-temporal multi-spectral satellite images from Sentinel-2 (Copernicus Open Access Hub, 2020), taken on 2/27/2020, 4/29/2020, 6/28/2020, 7/26/2020 and 8/20/2020, in 4 channels (R, G, B, NIR), with a maximal resolution 10 m; and an image from Landsat-8, taken on 4/04/2020 (Earth Explorer, 2020), in 8 channels, with a maximal resolution 15 m. Additionally, we used the data obtained during the field works in July-August 2020 in the territory of the Volga-Don irrigation system.
During the analysis of the images, a visual and automated decryption of the agricultural lands structure was carried out, using the spectral brightness of the fields with different crops, in 4 channels: 1 - blue (0.4-0.5 p,m), 2 - green (0.5-0.6 p,m), 3 - red (0.6- 0.7 |im), 4 - near infrared (0.7-0.95 |im). The vegetation indices (NDVI) were calculated, and various combinations of channels and image classification by the values of spectral brightness and NDVI were used.
Results and Discussion
The area that gets sown in the Volgograd Region covers more than 3 million ha, more than half of which is occupied with cereals and grain legumes (Local Agency ..., 2020). Wheat is the most widespread among winter cereals; barley, wheat, oats, corn for grain and ensilage are widespread among spring cereals; peas, chickpeas, sorghum and soybeans - among grain legumes (in limited amounts); mustard, sunflower, safflower, linseed - among industrial crops, and Lucerne with Sudan grass are among forage grasses.
In this region the winter crops of the dry steppe are usually sown in the I-II decade of September and harvested in the II decade of July. The spring crops are sown in different decades of April and harvested in July-August. Ensilage corn and Sudan grass are sown in May and harvested in August. Soybeans are sown from late April to early May and harvested in August (AGRIENE, 2020). According to this distribution, it becomes clear that a series of multi-temporal images should be used in order to identify the land structure in the most reliable way.
In 2020 the following crops were cultivated in the territory of the Volga-Don irrigation system:
1) winter cereals (non-irrigated);
2) spring cereals (non-irrigated)
3) ensilage corn (irrigated);
4) Sudan grass (irrigated and non-irrigated);
5) soybeans (irrigated).
Besides, there were leys and fallow fields.
To recognize the structure of agricultural lands using the satellite images, we had to take into account the following characteristics:
1) the leys on the images are patchy due to the surface erosion and the structure of the soil cover, both of which cause significant fluctuations of spectral brightness within one field;
2) not every field is irrigated, some of them are dry lands, therefore there is a differences in their soil moisture and spectral brightness;
3) some of the fields that are not currently irrigated still have the traces of buried temporary ditches, which affect the overall spectral brightness of the field;
4) the fields may vary in brightness due to agricultural industrial measures that were carried out on them, such as harrowing;
5) 2020 was arid, so the satellite images show its features on them.
Figure 1 shows fragments of images, taken in February, April, June, July and August 2020. The cultivated crops and the general structure of the fields are highlighted. For this we also used the materials obtained during the field studies. The images of such resolution allow us to identify winter
cereals and spring crops, leys and fallow. The further separation of spring crops is possible, but with a less probability. Moreover, the small fields and corner areas around the circular fields (0.2-1 ha) can be identified either poorly or not at all.
Further below the visual deciphering of the agricultural lands structure on the multi-temporal images is shown, by the example of the central part of the Volga-Don irrigation system.
Winter crops (1, 1a) are well-distinguished in February, March and April (Fig. 2 A, B, C). On the April image (Fig. 2 B) it is a perennial fallow (7).
On the June image (Fig. 2 C) these are ripening winter crops (1) and spring cereals (2), Sudan grass (3a), leys (6), fallow (7), but corn (4) and soybeans (5) are not that well distinguished.
On the July image (Fig. 2 D) there are harvested winter crops (1) and spring cereals (2), irrigated growing corn (4) and Sudan grass (3b), leys (6), but the color of non-irrigated Sudan grass (3a), soybeans (5) and fallow (7) barely differ from the leys.
n
a □ E
a a Q
Uptu- w A i ^^fc^jTjj a vv n
rn - rnSpfrjfù 1 * a u gl m 0
Jv □ HQft B JrVf' • '
rmmm ••"in 1 -■ r • E 23B3 "*. F 55
llSUM 111 ™ .¿ft / ^—"E
Fig. 2 A, B. Satellite image of the agricultural lands in the central part of the Volga-Don irrigation system (Volgograd Region), taken by Sentinel-2 on A) 2/27/2020, B) 3/28/2020. Legend: 1, la -winter cereals, 2 - spring cereals, 3 a - Sudan grass on the dry lands, 3b - irrigated Sudan grass, 4 -ensilage corn, 5 - soybeans, 6 - leys, 7 - perennial fallow.
□ ■ ° B □
B ,\ \ aB
n mm IS
a o a D □
□
LsLlBl
B 1 3 3 D
□ 3
3
□
EH 9
3 a
3 ES 3
□ □
3
in 3
IfffffiTI
□ IB 0 _ a
a o
U EO
□
3
fcvw m
a a
Fig. 2 C, D, E. Satellite image of the agricultural lands in the central part of the Volga-Don irrigation system (Volgograd Region), taken by Sentinel-2 on C) 4/29/2020, D) 5/27/2020, E) 8/20/2020. Legend: 1, la - winter cereals, 2 - spring cereals, 3a - Sudan grass on the dry lands, 3b - irrigated Sudan grass, 4 - ensilage corn, 5 - soybeans, 6 - leys, 7 - perennial fallow.
On the August image (Fig. 2 E) these are harvested winter crops (1) and spring cereals (2), soybeans (5), irrigated growing corn (4) and Sudan grass (3b), but leys (6) and fallow (7) are almost the same.
Each image does not provide much information about the agricultural lands structure, but, when combined, they give us a full understanding of the situation in this irrigation system.
After this, on the basis of the results of visual recognition, we created the training samples for an algorithm of automated sequential deciphering of the images taken at different times. We used such features as spectral brightness in 4 channels of Sentinel-2 and NDVI. All fields were digitized before we started. The training samples were marked on the images, taken on 2/27/2020, 4/29/2020, 6/28/2020 and 7/26/2020. Then we created an individual sample for each month and built a teaching model on the basis of Random Forest, using Python language. The scheme of the algorithm can be seen in Figure 3.
Fig. 3. Algorithm scheme of classification of the satellite imagery with agricultural lands, taken by Sentinel-2.
Random Forest is an ensemble of Decision Tree classifiers or a set of many decision trees. Each sample gets its own decision tree. Unlike an individual classifier, the ensemble tends to overfit less due to the usage of an average forecast for all its trees, resulting in satisfying image classification (Forests of Randomized Trees, 2020).
The results of the automated identification of agricultural lands structure on the satellite images for all dates mentioned above are shown below (Fig. 4).
On the February image (Fig. 4 A) the winter cereals on the big fields are well-distinguished (1), but the small (0.2-0.5 ha) fields (1a) are not. The rest of the fields are an open surface (8).
On the April image (Fig. 4 B) there are also fallows (7) aside from winter cereals (1). The rest of the open surface can be divided into different types by the level of their soil moisture (8, 8a).
On the June image (Fig. 4 C) the ripening winter crops can be identified (1, 1a) and spring cereals (2), non-irrigated Sudan grass (3 a), corn (4), soybeans (5), fallow (7) and leys, divided into several types (6, 6a, 6b, 6c).
D
IÜC33
A ■
CiSni
^
3 • :•• a
a '^v'. D3 a
3 (1
n 3
A
(5303
■ 4
a 5 a 6
6a
• Mi
■ 6c
■I »thrr
Fig. 4 A, B, C. Results of the automated series classification of the satellite images from Sentinel-2 that was based on Random Forest, and identification of the agricultural lands in the central part of the Volga-Don irrigation system (Volgograd Region) for A) 2/27/2020, B) 3/28/2020, C) 4/29/2020. Legend: 1, 1a - winter cereals, 2, 2a - spring cereals, 3a - Sudan grass on a dry land field, 3b - irrigated Sudan grass, 4 - ensilage corn, 6, 6a, 6b, 6c - leys, 7 - perennial fallow, 8 -open field surface.
Fig. 4 D, E. Results of the automated series classification of the satellite images from Sentinel-2 that was based on Random Forest, and identification of the agricultural lands in the central part of the Volga-Don irrigation system (Volgograd Region) for D) 5/27/2020, E) 8/20/2020. Legend: 1, 1a - winter cereals, 2, 2a - spring cereals, 3a - Sudan grass on a dry land field, 3b - irrigated Sudan grass, 4 - ensilage corn, 6, 6a, 6b, 6c - leys, 7 - perennial fallow, 8 - open field surface.
On the July image (Fig. 4 D) aside from the mentioned crops there is also irrigated Sudan grass (3b), while the leys are divided into two types (6, 6a).
On the August image (Fig. 4 E) there are harvested winter crops (1, 1a) and spring cereals (2), soybeans (5), growing corn (4) and Sudan grass (3b), fallow (7) and leys of two types (6, 6a).
An automated series classification of satellite imagery showed that identification of the agricultural lands structure is possible. However, due to small fields being undistinguished sometimes and ley fields dividing into several classes by the level of their soil moisture, the final stage map compilation requires us to combine those fields separately.
According to the results of this classification, we created a generalized error matrix that can be found in Figure 5. Along its diagonal are the values that indicate the share of correctly classified pixels. Anything that is outside the diagonal is marked as the I and II type errors. The average share of correctly classified pixels for all fields was 75%.
Therefore, with the use of multi-temporal satellite imagery for highlighting of the lands structure, it is possible to classify the image both visually and automatically, with an accuracy of 75%.
When conducting the soil surveys and choosing the key sites for them, it is important to focus on the perennial fallow on previously irrigated lands, which usually the most problematic fields turn into. A special attention is required when studying a young fallow under 5 years old, which could probably form due to the soil degradation.
1 2 34 3b 4 J 6 7 10 otfwf
1 0.76497 0.20595 0.03059 0 00485 0024989 0 004363 0011981 0051899 0.100128
z 0.07622 0-72369 0.01264 0-00194 0-006181 0 001227 0006879 0 001006 0.018218
3a 0 00054 0 00019 0,76726 0 0 010243 0 14E-05 0 000567 4.37E-05 0.001278
lb 1.4i-0J O 0.00064 0-66731 0.024283 0 0.00331 0-001417 2.19t -OS 0.00200S
4 0.00)48 0 00287 0.03249 0.1096 0.79011 025 001807 0.010356 0001378 0.016272
S 0 0 0.00023 0 0 006181 0 73684 4 2E-05 S67T-C5 0 0,00049«
6 0.03134 0 02266 0.10013 01581 0 063929 0 0-81623 0.028629 0.025217 0.068253
7 0.00387 0-0041B 0.02099 0.0291 0.030022 0 0.01614 0.78119 0.00573 0,166173
10 0.01254 0.00217 0.0019 0.00097 0.012539 0 0.03491 0-012604 0-763051 0.101MJ
o&ief 0.10637 0 03729 0.03312 0 02513 0-031523 0-01316 0-05539 0 146321 0.151652 0 i 75976
Fig. 5. Generalized error matrix.
There is also a large number of non-irrigated fields in the territory of the Volga-Don irrigation system, which may be a sign of cost savings for irrigation, as well as the patchy and complex soil cover, a significant amount of saline and solonetzic soils that force the farms to leave these fields without any irrigation.
To separate irrigated and non-irrigated fields, to identify and divide the fallows under and over 5 years old, we select empirically and use a combination of different channels for the multispectral satellite images (Fig. 6, 7).
When identifying a perennial fallow on the image taken by Landsat-8 on 4/04/2020, using a combination of channels 7, 5 and 4 (short-wave infrared, near infrared, red), a younger deposit (under 5 years old) is highlighted with saturated pink, while an older one is highlighted with pale pink (Fig. 6 A). The reliability of this identification can be proved by comparing the new image with the older one, taken 5 years ago and shown below (Fig. 6 B).
The irrigated areas are marked with violet on a satellite image from Sentinel-2, taken on 7/27/2020, in a combination of the channels 2, 3 and 4 (green, red, near infrared; Fig. 7).
Based on the final results, we created a map of the agricultural lands structure for the central part of the Volga-Don irrigation system. It shows the fields with different crops, whether they are irrigated or not, as well as it shows the leys, uneven-aged fallow and the areas of all fields (Fig. 8).
We determined that 35% of the total area of the studied agricultural fields (8988.9 ha) was cultivated, 36.8% was left under the leys, and 28.6% turned into a fallow, which is a significant share. But only 3.5% of the cultivated fields were irrigated, which is unacceptably low. The reason for poor lands development can be economic problems, as well as degraded and complex soil cover.
Fig. 6. A) perennial fallow on the satellite image taken on 4/04/2020 by Landsat-8, in a combination of channels 7, 5 and 4 (short-wave infrared, near infrared, red); B) fields on the satellite image from Google.Earth, taken on 8/05/2015, than turned into a perennial fallow by 2020. Legend. A) 1 - fallow under 5 years old, 2 - fallow over 5 years old; B) 1 - harvested field, 2 -arable land, 3 - ley field, 4 - perennial fallow.
Fig. 7. Violet (1) marks the irrigated areas on the image taken on 7/27/2020 by Sentinel-2, in a combination of the channels 2, 3 and 4 (green, red, near infrared).
The total area of 130 fields was 8988.9 ha. The winter cereals occupied 2026.2 ha (22.5%), spring crops - 493.0 ha (5.5%), irrigated Sudan grass - 44.2 ha (0.5 %), non-irrigated - 300.5 ha (3.3%), irrigated ensilage corn - 255.0 ha (2.8%), irrigated soybeans - 3.4 ha (0.05%), leys - 3304.3 ha (36.8%), fallow under 5 years old - 1205.8 ha (13.4%), over 5 years old - 1356.5 ha (15.1%).
This means that 35% of the total fields were cultivated, 36.8% were left under the leys, and 28.6% became a fallow, which is a significant amount. But only 3.5% of the cultivated fields were irrigated, which is unacceptably low. The reason for this could be economic problem of the farms and degraded, complex soil cover that requires an expensive amelioration.
We assessed the cultivated crops condition, using NDVI (Table 1).
When the vegetation cover is sparse, the spectrum depends mostly on soil, and soils background affects NDVI severely. Due to the crops diversity, patchiness of the soil cover and dry weather of 2020, the NDVI values range significantly between different types of crops in the territory of the Volga-Don irrigation system.
The NDVI values are quite even (0.13-0.17) for the ley fields during the entire period of our study, which matches the values of the open soil surface.
Winter cereals reach their highest index NDVI in April (0.72-0.73), ripe by June (0.2-0.21) and are harvested in July. The index for corn, soybeans and irrigated Sudan grass is at its maximum in August (0.60-0.88), when the harvest begins. The NDVI values for fallow are in-between leys and growing crops (0.20-0.40) due to the weeds, the index of which reaches its highest in June (0.380.48), but never exceeds 0.5.
The spring cereals were noted to have a special way of development. We analyzed them on the images separately, because the time intervals, suggested in Table 1, could not cover their short growing season. The cereals were growing most actively in the end of May, and reached maturity at the beginning of July and were harvested. The maximum NDVI (0.52-0.54) indicated they were not growing sufficiently enough, which could be due to the dry year, and, perhaps, it was the exact reason for an early harvest (Table 2).
If the range of NDVI values on a field with only one type of crop on the active stage of growing differs by 0.1 and more (Sudan grass on dry land, soybeans), it is a sign of its uneven development, which can be due to the patchy soil cover and, therefore, requires a special treatment when the key sites are selected.
No. Code Subject Area, ha No. Code Subject Area, ha
29 7 Fallow under 5 years 229,7 49 7 Fallow over 5 years 19,8
30 7 Fallow under 5 years 233,4 50 7 Fallow over 5 years 14,5
31 6 Leys 52,3 51 1 Winter cereals 111,1
32 3 Sudan grass on dry land 167,3 52 1 Winter cereals 19,1
33 6 Leys 193,5 53 7 Fallow over 5 years 123,3
34 3 Sudan grass on dry land 59,2 54 1 Winter cereals 140,5
35 6 Leys 59,8 55 7 Fallow over 5 years 247,8
36 2 Spnng cereals 161,6 56 1 Wnter cereals 337,4
37 1 Winter cereals 128,2 57 6 Leys 230,8
38 7 Fallow under 5 years 87,8 58 7 Fallow over 5 years 86,7
39 6 Leys 172,5 59 7 Fallow over 5 years 89,5
40 7 Fallow under 5 years 41,4 60 6 Leys 12,5
41 7 Fallow over 5 years 7,7 61 1 Wnter cereals 40,8
42 7 Fallow over 5 years 12,7 62 6 Leys 74,2
43 7 Fallow over 5 years 17,1 63 1 Winter cereals 60,0
44 7 Fallow over 5 years 17,1 64 6 Leys 16,5
45 7 Fallow over 5 years 16,4 65 6 Leys 117,2
46 7 Fallow over 5 years 25,1 66 1 Wnter cereals 105,4
47 7 Fallow over 5 years 27,8 67 4 Irrigated ensilage corn 65,6
48 7 Fallow over 5 years 17,5 68 4 Imgated ensilage corn 44,2
Fig. 8. Structure of the agricultural lands in the central part of the Volga-Don irrigation system (Volgograd Region) in 2020 and the area of each field. Notes: 1 - winter cereals, 2 - spring cereals, 3 a - Sudan grass on dry lands, 3b - irrigated Sudan grass, 4 - irrigated ensilage corn, 5 - irrigated soybeans, 6 - leys, 7a - fallow under 5 years old, 76 - fallow over 5 years old, 63-74 - field number.
To make a general representation of the way the NDVI values distributed over the territory of the central part of the Volga-Don system, we classified the satellite image from Sentinel-2, taken in July (Fig. 9). It can be seen on the Figure 9 that there are fields with harvested crops and leys at NDVI<0.2, fields with fallow and Sudan grass on dry land at 0.2-0.4, and fields with variously growing irrigated crops (corn, soybeans, Sudan grass) at 0.4-1, as well as garden plots in the villages and heavily vegetating plants in the gullies.
Therefore, the analysis of the NDVI values makes it possible to separate the growing crops, leys and fallow, to determine the state of the crop and use its uneven development to determine the areas with the highest patchiness of the soil cover.
Table 1. NDVI values for the agricultural lands of the Volga-Don irrigation system in February-August 2020.
No. crop type February April June July August
1 winter cereals 0.62-0.70 0.72-0.73 0.20-0.21 harvested harvested
3a Sudan grass on the dry land 0.64-0.76 0.28-0.30 0.26-0.28
3b irrigated Sudan grass 0.60-0.66 0.60-0.80 0.60-0.85
4 ensilage corn 0.69-0.71 0.83-0.86 0.80-0.85
5 soybeans 0.37-0.47 0.47-0.50 0.79-0.88
6 leys 0.15-0.17 0.14-0.19 0.16-0.18 0.14-0.15 0.13-0.15
7 fallow 0.34-0.40 0.38-0.48 0.24-0.31 0.20-0.21
Table 2. NDVI values for the spring cereals in the territory of the Volga-Don irrigation system in May-August 2020.
No. crop type May June early July late July August
2 spring cereals 0.52-0.54 0.23-0.25 0.14-0.15 harvested harvested
Fig. 9. Classification by the NDVI values of the satellite image from Sentinel-2, taken on 7/27/2020 in the central part of the Volga-Don irrigation system.
Then, based on the obtained results, we selected several key sites for further study of irrigated soils (Fig. 10). The first site was a field with Sudan grass on dry soils, with NDVI less than 0.4 during the active phase of the growing season, which was a sign of soil degradation that were reducing the vegetation activity of the crop (most likely, it was alkaline soil). The second one was selected due to its patchy ley field, which indicated the erosional soil degradation. The third one was on a field with buried temporary ditches, which implied a strong anthropogenic impact on soils and a diversity of soil formations. The fourth one was with a fallow over 5 years old, which indicated a complex soil cover and made it necessary to leave the field without cultivation and irrigation for a long period of time. The fifth one was an irrigated field with a successfully growing crop (silage corn), where the NDVI was more than 0.8.
Fig. 10. Selected key sites (1-5) for a soil survey, marked on a satellite image from Sentinel-2, 8/27/2020.
Conclusions
This study aimed to identify the possibility to recognize the structure of agricultural lands (fields with various crops, leys and fallows), using the multispectral high resolution satellite imagery (10 m) from Sentinel-2. It took place in the area of Volga-Don irrigation system that was located in the dry steppe solonetz zone of Volgograd Region. The idea was to select some key areas for soil survey afterwards, while taking into account the state of crops and fields. For this we analyzed the images, taken in the said area in February-August 2020, and used materials from field surveys that were carried out in August-September 2020. The structure recognition was made visually and automatically.
In the process we discovered that after a sequential visual analysis of multi-temporal multispectral satellite images was carried out, the winter crops stood out the most in February, March and April. On the April image it was the perennial fallow that was well defined; in June it was the ripening winter and spring crops, Sudan grass on dry land, leys and fallow, but irrigated corn and soybeans were barely recognizable; in July the harvested winter and spring crops, corn and leys were clearly visible, but the color of Sudan grass on dry land, soybeans and fallow hardly differed from the leys; in August the harvested winter and spring crops, soybeans and corn were
well-defined, but leys and fallow did not differ too much. The satellite images that had been taken at different times, made it possible for us to mark all cultivated fields on the irrigation system.
We used a Random Forest method, based on an ensemble of Decision Tree classifiers, to automatically classify satellite imagery and highlight the structure of agricultural lands. In this method we used training samples that were based on visual interpretation of images, as well as took into account such features as spectral brightness in 4 channels of Sentinel-2, and NDVI values. The images that were processed had been taken in February-August 2020. According to the results, the structure was the clearest for June and July. We found out that small fields (with an area of 0.20.5 ha) could not be identified separately, because the merged with the surrounding background; while the ley fields could divide into 2 or 3 classes, depending on the month when the image was taken (and the difference in soil moisture), and therefore we had to combine them at the final stage of the processing.
The results helped us to comply a generalized error matrix, which showed that the average share of correctly classified pixels was 75% for the analyzed agricultural lands, therefore, this method of image processing can be considered acceptable.
The combination of different channels allowed us to separate the irrigated and non-irrigated fields on the irrigation system from the fallows of different ages.
After the work was finished, we created a map of the agricultural lands structure for the central part of the Volga-Don irrigation system. It shows the fields with different crops, whether they are irrigated or not, as well as it shows the leys, uneven-aged fallow and the areas of all fields.
We determined that 35% of the total area of the studied agricultural fields (8988.9 ha) was cultivated, 36.8% was left under the leys, and 28.6% turned into a fallow, which is a significant share. But only 3.5% of the cultivated fields were irrigated, which is unacceptably low. The reason for poor lands development can be economic problem, as well as degraded and complex soil cover.
We calculated the NDVI values to reflect the crops state at the different stages of their growth. The analyzed material made it possible to select key sites for the further soil survey, since the state of crops, among many other factors, reflects the structure and degradation of soil cover, and helps to select correctly the said sites to study and develop the required ameliorative measures for irrigated soils fertility improvement in the future.
REFERENCES REFERENCES
1. AGRIENE, Agrarian Internet Encyclopedia 1. АГРИЭН. Аграрная интернет-[AGRIEN, Agrarnaya internet- энциклопедия. 2020 [Available at entsiklopediya]. 2020, available at http://www.agrien.ru (Date of Accesse http://www.agrien.ru, (accessed 11.05.2021)].
11/05/2021). 2. Андронников В.Л. 1979.
2. Andronnikov VL. Aerospace methods for Аэрокосмические методы изучения почв. soils study [Aerokosmicheskiye metody М.: Колос. 280 с.
izucheniya pochv]. Moscow: Kolos, 3. Антипов-Каратаев И.Н. 1953.
1979:280. Мелиорация солонцов в СССР. М.: Изд-
3. Antipov-Karataev IN. Amelioration of во АН СССР. 563 с.
solonetzes in the USSR [Melioratsiya 4. Антонова С.Ю., Кравцова В.И. 1976.
solontsov v SSSR]. Moscow: AN SSSR, Опыт использования космических
1953:563. многозональных снимков для изучения
4. Antonova SYu., Kravtsova VI. Experience засоления территории // Исследования in the use of multi-spectral space images to природной среды космическими study the salinization of the territory [Opyt средствами. М.: ВИНИТИ. С. 131-139. ispol'zovaniya kosmicheskikh 5. Барановская А.В., Азовцев В.И. 1981.
mnogozonal'nykh snimkov dlya izucheniya zasoleniya territorii] Study of the natural environment with the usage of satellite images [Issledovaniya prirodnoy sredy kosmicheskimi sredstvami]. Moscow: VINITI, 1976:131-139.
5. Baranovskaya AV., Azovtsev VI. Influence of irrigation on the migration of carbonates in the soils of the Volga region [Vliyaniye orosheniya na migratsiyu karbonatov v pochvakh Povolzh'ya] Pochvovedenie. 1981;10:17-26.
6. Bezdnina SYa. Irrigation water quality: principles and methods for evaluation [Kachestvo vody dlya orosheniya: Printsipy i metody otsenki]. Moscow: ROMA, 1997:185.
7. Bolshakov AF. Water regime of powerful chernozems of the Central Russian Upland [Vodnyy rezhim moshchnykh chernozemov Srednerusskoy vozvyshennosti]. Moscow: AN SSSR, 1961:200.
8. VEGA-RRO, Vegetation analysis satellite service [VEGA-RRO, Sputnikovyy servis analiza vegetatsii]. 2020, available at http://pro-vega.ru, (accessed 11/05/2021).
9. Vyshivkin DD. Geographic interpretation of satellite images for the purpose of thematic mapping of saline lands [Geograficheskaya interpretatsiya kosmicheskikh snimkov v tselyakh tematicheskogo kartirovaniya zasolennykh zemel'] Research of the Natural Environment by Space Means [Issledovaniya prirodnoy sredy kosmicheskimi sredstvami]. Moscow: VINITI, 1975;4:34-37.
10. Gorokhova IN, Avdeeva TN, Pankova EI, Prokopyeva KO. Soil and agrochemical characteristics of the Svetloyar irrigated site in Volgograd Oblast Arid Ecosystems. 2019;(9)1:41-50.
11. Gorokhova IN, Pankova EI. The nature of the patchiness of irrigated soils in the dry steppe zone (on example of the Svetloyarskaya irrigation system) Arid Ecosystems. 2017;(7)3:161-170.
12. Gorokhova IN, Pankova EI, Chursin IN. Development of Methodological
Влияние орошения на миграцию карбонатов в почвах Поволжья // Почвоведение. № 10. С. 17-26.
6. Безднина С.Я. 1997. Качество воды для орошения: принципы и методы оценки. М.: РОМА. 185 с.
7. Большаков А.Ф. 1961. Водный режим мощных черноземов Среднерусской возвышенности. М.: Изд-во АН СССР. 200 с.
8. ВЕГА-FRO. Спутниковый сервис анализа вегетации. 2020 [Available at http://pro-vega.ru (Date of Accesse 11.05.2021)].
9. Вышивкин Д.Д. 1975. Географическая интерпретация космических снимков в целях тематического картирования засоленных земель // Исследования природной среды космическими средствами. М.: ВИНИТИ. Т. 4. С. 34-37.
10. Горохова И.Н., Авдеева Т.Н., Панкова Е.И., Прокопьева К.О. 2019. Почвенно-агрохимическая характеристика Светлоярского орошаемого участка в Волгоградской области // Аридные экосистемы. Т. 25. № 1. С. 48-59.
11. Горохова И.Н., Панкова Е.И. 2017. Природа пятнистости орошаемых почв сухостепной зоны (на примере Светлоярской оросительной системы) // Аридные экосистемы. Т. 23. № 3 (72). С. 44-54.
12. Горохова И.Н., Панкова Е.И., Чурсин И.Н. 2020. Методические подходы к разработке мониторинга орошаемых земель юга Европейской части России на основе материалов космической съемки // Аридные экосистемы. Т. 26. № 1 (82). С. 84-93.
13. Горохова И.Н., Хитров Н.Б., Кравченко Е.П. 2020. Изменение засоленности орошаемых почв участка Червленое за четверть века (Волгоградская область) // Почвоведение. № 4. С. 463-472.
14. Горохова И.Н., Хитров Н.Б., Панкова Е.И., Прокопьева К.О. 2018а. Засоленность почв Светлоярского орошаемого массива в Волгоградской области в 2010-х годах // Бюллетень
Approaches for the Assessment of the Status of Irrigated Lands in the South of the European Part of Russia Based on Space-Surveying Materials Arid Ecosystems. 2020;(l0)1:70-78.
13. Gorokhova IN, Khitrov NB, Kravchenko EP. Changes in soil salinity at the Chervlenoe irrigation massive (Volgograd Oblast) in a quarter of century Soil Science. 2020;4:463-472.
14. Gorokhova IN, Khitrov NB, Pankova EI, Prokopyeva KO. Soil salinity of the Svetloyarsk irrigated massif in the Volgograd region in the 2010s [Zasolennost' pochv Svetloyarskogo trebuyemogo kachestva v Volgogradskoy oblasti v 2010-kh godakh] Bulletin of the V. V. Dokuchaev Soil Institute. 2018a;(93):75-93.
15. Gorokhova IN, Khitrov NB, Prokopyeva KO, Harlanov VA. Soil Cover of the Svetloyarsk irrigation system after 50 years of reclamation practices Eurasian Soil Science. 2018b;(51)8:1-11.
16. Degtyareva ET, Zhulidova AN. Soils of the Volgograd region [Pochvy Volgogradskoy oblasti]. Volgograd: Nizhne-Volzhsky Publishing House, 1970:319.
17. Dedova EB. Zonal scale for assessing the quality of irrigation waters in the Republic of Kalmykia // Synergy [Zonal'naya shkala otsenki kachestva polivnykh vod respubliki Kalmykiya]. 2018;1:88-95, available at https://vepi.ru/wp-
content/uploads/2018/10/Sinergiya-2018-1.pdf (accessed 17/05/2020).
18. Zaidelman FR. Environmental protection of reclaimed soils and agricultural landscapes [Ekologicheskaya zashchita melioriruyemykh pochv i agrolandshaftov] Pochvovedenie. 1993;1:5-12.
19. Zimovets BA. Ecology and soil reclamation in the dry steppe zone [Ekologiya i melioratsiya pochv sukhostepnoy zony]. Moscow: V.V. Dokuchaev Soil Institute, 1981:248.
20. Zinchenko EV, Gorokhova IN, Kruglyakova NG, Khitrov NB. The current state of irrigated soils in the south of the Volga
Почвенного института им. В. В. Докучаева. Вып. 93. С. 75-93.
15. Горохова И.Н., Хитров Н.Б., Прокопьева К.О., Харланов В.А. 2018б. Почвенный покров Светлоярской оросительной системы через полвека мелиоративных воздействий // Почвоведение. № 8. С. 118.
16.Дегтярева Е.Т., Жулидова А.Н. 1970. Почвы Волгоградской области. Волгоград: Нижне-Волжское книжное изд-во. 319 с.
17. Дедова Э.Б. 2018. Зональная шкала оценки качества поливных вод республика Калмыкия // Синергия. № 1. С. 88-95 [Available at https://vepi.ru/wp-content/uploads/2018/10/Sinergiya-2018-1.pdf (Date of Access 17.05.2020)].
18. Зайдельман Ф.Р. 1993. Экологическая защита мелиорируемых почв и агроландшафтов // Почвоведение. № 1. С. 5-12.
19. Зимовец Б.А. 1981. Экология и мелиорация почв сухостепной зоны. М.: Почвенный институт им. В.В. Докучаева. 248 с.
20. Зинченко Е.В., Горохова И.Н., Круглякова Н.Г., Хитров Н.Б. 2020. Современное состояние орошаемых почв юга Приволжской возвышенности // Бюллетень Почвенного института имени В.В. Докучаева. Вып. 104. М.: РАСХН. С. 68-109.
21. Иванова Е.И. 1928. Очерк почв южной части Подуральского плато и прилегающих районов Прикаспийской низменности. Л.: Изд-во АН СССР. 320 с.
22. Ковда В.А. 1946-1947. Происхождение и режим засоленных почв. М.-Л.: Изд-во АН СССР. Т. 1, 575 с.; Т. 2, 375 с.
23. Ковда В.А. 1937. Солончаки и солонцы М.-Л.: Изд-во АН СССР. 246 с.
24. Козловский Ф.И. 1991. Современные естественные и антропогенные процессы эволюции почв. М.: Наука. 198 с.
25. Конюшкова М.В. 2014. Цифровое картографирование почв солонцовых комплексов Северного Прикаспия. М.:
Upland [Sovremennoye sostoyaniye oroshayemykh pochv yuga Privolzhskoy vozvyshennosti] Bulletin of the V.V. Dokuchaev Soil Institute. 2020;(104):68-109.
21. Ivanova EI. Review of the soils of the southern part of the Pre-Ural Plateau and adjacent areas of the Caspian Lowland [Ocherk pochv yuzhnoy chasti Podural'skogo plato i prilegayushchikh otritsatel'nykh Prikaspiyskoy nizmennosti]. Leningrad: AN SSSR. 1928:320.
22. Kovda VA. Origin and regime of saline soils [Proiskhozhdeniye i rezhim zasolennykh pochv]. Moscow-Leningrad: AN SSSR, 1946-1947;(1):575; (2):375.
23. Kovda VA. Salines and solonetzes [Solonchaki i solontsy]. Moscow-Leningrad: AN SSSR, 1937:246.
24. Kozlovsky FI. Modern natural and anthropogenic processes of soil evolution [Sovremennyye yestestvennyye i antropogennyye protsessy evolyutsii pochv]. Moscow: Science, 1991:198.
25. Konyushkova MV. Digital mapping of the soils of the solonetz complexes of the Northern Caspian region [Tsifrovoye kartografirovaniye pochv solontsovykh kompleksov Severnogo Prikaspiya]. Moscow: KMK Scientific Publishing Association, 2014:316.
26. Koroleva PV, Rukhovich DI, Rukhovich AD, Rukhovich DD, Kulyanitsa AL, Trubnikov AV, Kalinina NV, Simakova MS. Location of the open soil surface and soil line in the RED-NIR spectral space [Mestopolozheniye otkrytoy poverkhnosti pochvy i linii v spektral'nom prostranstve RED-NIR] Pochvovedenie. 2017;12:1435-1446.
27. Kuznetsov MS, Grigoriev VYa, Khan KYu. Irrigational soil erosion and its prevention during sprinkler irrigation [Irrigatsionnaya eroziya pochv i yeye preduprezhdeniye pri polivakh dozhdevaniyem]. Moscow: Science, 1990:120.
28. Lyubimova IN, Degtyareva ET. Changes in
Товарищество научных изданий КМК. 316 с.
26. Королева П.В., Рухович Д.И., Рухович А.Д., Рухович Д.Д., Куляница А.Л., Трубников А.В., Калинина Н.В., Симакова М.С. 2017. Местоположение открытой поверхности почвы и линии почвы в спектральном пространстве RED-NIR // Почвоведение. № 12. С. 1435-1446.
27. Кузнецов М.С., Григорьев В.Я., Хан К.Ю. 1990. Ирригационная эрозия почв и ее предупреждение при поливах дождеванием. М.: Наука. 120 с.
28. Любимова И.Н., Дегтярева Е.Т. 2000. Изменение карбонатного профиля почв солонцовых комплексов при агрогенном воздействии // Почвоведение. № 7. С. 855-860.
29. Любимова И.Н., Новикова А.Ф. 2016. Влияние различных антропогенных воздействий на изменение почв солонцовых комплексов сухостепной зоны // Почвоведение. № 5. С. 633-643.
30. Мамедов Э. А. 1985. Изучение засоленных земель и солончаков с помощью космических методов // Исследование Земли из космоса. № 1. С. 60-61.
31. Минашина Н. Г. 1978. Мелиорация засоленных почв. М.: Колос. 263 с.
32. Панкова Е.И., Горохова И.Н., Конюшкова М.В. 2014. Развитие методов оценки и дистанционного мониторинга засоления орошаемых и целинных почв на территории аридных и семиаридных регионов // Экологические системы и приборы. № 10. С. 3-9
33. Панкова Е.И., Мазиков В.М. 1975. Методические вопросы использования аэрофотоснимков для характеристики засоления почв. Почвенно-мелиоративные процессы в районах нового орошения // Научные труды Почвенного института им. В.В. Докучаева. М. С. 97-111.
34. Панкова Е.И., Соловьев Д.А. 1993. Дистанционный мониторинг засоления орошаемых почв. М.: Почвенный
the Carbonate Distribution in the Soils of Solonetzic Complexes at Agrogenic Impact Eurasian Soil Science. 2000;(33)7:746-751.
29. Lyubimova IN, Novikova AF. Changes in the Properties of Solonetzic Soil Complexes in the Dry Steppe Zone under Anthropogenic Impacts Eurasian Soil Science. 2016;(49)5, 581-590.
30. Mamedov EA. Study of saline soils and solonetzes using satellite images [Izucheniye zasolennykh zemel' i solonchakov s pomoshch'yu kosmicheskikh metodov] Earth Research from Space [Issledovaniye Zemli iz kosmosa]. 1985;1:60-61.
31. Minashina NG. Amelioration of saline soils [Melioratsiya zasolennykh pochv]. Moscow: Kolos, 1978:263.
32. Pankova EI, Gorokhova IN, Konyushkova MV. Development of methods for assessment and remote monitoring of salinization of irrigated and virgin soils on the territory of arid and semi-arid regions [Razvitiye metodov otsenki i distantsionnogo monitoringa zasoleniya oroshayemykh i tselinnykh pochv na territorii aridnykh i semiaridnykh regionov] Ecological Systems and Devices [Ekologicheskiye sistemy i pribory]. 2014;10:3-9
33. Pankova EI, Mazikov VM. Methodological issues of using aerial photographs to characterize soil salinity. Soil reclamation processes in areas of new irrigation [Metodicheskiye voprosy ispol'zovaniya aerofotosnimkov dlya kharakteristik zasoleniya pochv. Pochvenno-meliorativnyye protsessy v rayonakh novogo orosheniya] Proceedings of the V.V. Dokuchaev Soil Institute [Nauchnyye trudy Pochvennogo instituta im. V.V. Dokuchayeva]. Moscow, 1975:97-111.
34. Pankova EI, Soloviev DA. Remote monitoring of salinity in irrigated soils [Distantsionnyy monitoring zasoleniya oroshayemykh pochv]. Moscow: V.V. Dokuchaev Soil Institute, 1993:191.
35. Prikhodko VE. Irrigated steppe soils: functioning, ecology, productivity
институт им. В.В. Докучаева. 191 с.
35. Приходько В.Е. 1996. Орошаемые степные почвы: функционирование, экология, продуктивность. М.: Интеллект. 179 с.
36. Роде А.А. 1947. Почвообразовательный процесс и эволюция почв. М.: Географгиз. 142 с.
37. Розанов Б. Г. 1975. Генетическая морфология почв. М.: Изд-во МГУ. 293 с.
38. Рухович Д.И. 2009. Многолетняя динамика засоления орошаемых почв центральной части Голодной степи и методы ее выявления. Дисс. ... к.б.н. М. 25 с.
39. Савин И.Ю., Отаров А., Жогалев А.В., Ибраева М.А., Дуйсеков С. 2014. Выявление многолетних изменений площади засоленных почв Шаульдерского орошаемого массива по космическим снимкам Landsat // Бюллетень почвенного института им. В.В. Докучаева. Вып. 74. М.: РАСХН. С. 49-63.
40. Сиземская М.Л. 2013. Современная природно-антропогенная трансформация почв полупустыни Северного Прикаспия. М.: Товарищество научных изданий КМК. С. 276.
41. Симакова М.С. 1959. Методика картирования почв Прикаспийской низменности по материалам аэрофотосъемки // Почвенно-географические исследования и использование аэрофотоснимков в картографировании почв. М.: Изд-во АН СССР. С. 283-357.
42. Территориальный орган Федеральной службы государственной статистики по Волгоградской области. 2020 [Available at https://volgastat.gks.ru (Date of Access 17.05.2020)].
43. Харитонов В.А. 1982. Методика и результаты определения степени засоленности орошаемых земель Ферганской области по космическим снимкам // Рациональные методы и средства в мелиорации. М.: ВНИИГИМ.
[Oroshayemyye stepnyye pochvy: funktsionirovaniye, ekologiya,
produktivnost']. Moscow: Intellect, 1996:179.
36. Rode AA. Soil formation and evolution [Pochvoobrazovatel'nyy protsess i evolyutsiya pochv]. Moscow: Geografgiz, 1947:142.
37. Rozanov BG. Genetic morphology of soils [Geneticheskaya morfologiya pochv]. Moscow: Publishing House of Moscow State University, 1975:293.
38. Rukhovich DI. Long-term dynamics of irrigated soils salinization in the central part of Mirzacho'l and methods of its identification [Mnogoletnyaya dinamika zasoleniya oroshayemykh pochv chasti Golodnoy stepi i metody yeye tekushchey]. PhD dissertation, Moscow, 2009:25.
39. Savin IYu, Otarov A, Zhogalev AV, Ibrayeva MA, Duisekov S. Identification of long-term changes in the area of saline soils of the Shaulder irrigated massif based on Landsat satellite images [Vyyavleniye mnogoletnikh izmeneniy ploshchadi zasolennykh pochv Shaul'derskogo oroshennogo massiva po kosmicheskim snimkam Landsat] Bulletin of the V.V. Dokuchaev Soil Institute. 2014;(74):49-63.
40. Sizemskaya ML. Modern natural and anthropogenic transformation of soils in the semi-desert of the Northern Caspian region [Sovremennaya prirodno-antropogennaya transformatsiya pochv polupustyni Severnogo Prikaspiya]. Moscow: KMK Scientific Publishing Association, 2013:276.
41. Simakova MS. Methods for mapping the soils of the Caspian lowland on the basis of aerial photography [Metodika kartirovaniya pochv Prikaspiyskoy nizmennosti po materialam aerofotos 'yemki] Soil-geographical research and the use of aerial photographs in soil mapping [Pochvenno-geograficheskiye issledovaniya i ispol'zovaniye aerofotosnimkov v kartografirovanii pochv]. Moscow: AN SSSR, 1959:283-357.
42. Local Agency of the Federal Service ^a
С. 107-115.
44. Хитров Н.Б., Чечуева О.А., Никитина Н.С., Гришина Р.В. 1994. Влияние ионно-солевого состава почвенного раствора на разрушение микроагрегатов карбонатного чернозема // Почвоведение. № 1. С. 51-60.
45.Abbas A., Khan S., Hussain N., Hanjra M.A., Akbar S. 2013. Characterizing soil salinity in irrigated agriculture using a remote sensing approach // Physics and Chemistry of the Earth. Part A/B/C/. Vol. 55-57. P. 1-8.
46. Allbed A., Kumar L., Aldakheel Y.Y. 2014. Assessing soil salinity using soil salinity and vegetation indices derived from IKONOS high-spatial resolution imageries: Applications in a date palm dominated region // Geoderma. Vol. 230-231. P. 1-8.
47. Bahmaei A., Albaji M., Naseri A.A., Varjavand P. 2020 . Effect of irrigation type and interval on soil salinity in clay soils in Ahvaz, Iran // Arabian Journal of Geosciences. No. 13. P. 1151.
48. Chi Y., Sun J., Liu W., Wang J., Zhao M. 2019. Mapping coastal wetland soil salinity in different seasons using an improved comprehensive land surface factor system // Ecological Indicators. Vol. 107. P. 105517.
49. Copernicus Open Access Hub. 2020 [Available at https://scihub.copernicus.eu/dhus/#/home (Date of Access 10/04/2021)].
50. Crippen R.E. 1990. Calculating the Vegetation Index Faster // Remote Sensing of Environment. Vol. 34. P. 71-73.
51. Dwivedi R.S. 1996. Monitoring of salt-affected soils of the Indo-Gangetic alluvial plains using principal component analysis // International Journal of Remote Sensing. Vol. 17. No. 10. P. 1907-1914.
52. Earth Explorer. 2020 [Available at https://earthexplorer.usgs.gov/ (Date of Access 10.04.2021)].
53. Forests of Randomized Trees. 2020 [Available at https://scikit-learn.org/stable/modules/ensemble.html#for est (Date of Access 15.04.2021)].
54. Gorji T., Sertel E., Tanik A. 2017.
State Statistics for the Volgograd Region [Territorial'nyy organ Federal'noy sluzhby gosudarstvennoy statistiki po
Volgogradskoy oblasti]. 2020, available at https://volgastat.gks.ru, (accessed
17/05/2020).
43. Kharitonov VA. Methods and results of determining the degree of salinity of irrigated lands in the Fergana region from satellite images [Metodika i rezul'taty opredeleniya stepeni zasolennosti oroshayemykh zemel' Ferganskoy oblasti po kosmicheskim snimkam] Rational methods and means in land reclamation [Ratsional'nyye metody i sredstva v melioratsii]. Moscow: VNIIGIM, 1982:107115.
44. Khitrov NB, Chechueva OA, Nikitina NS, Grishina RV. Influence of the ion-salt composition of the soil solution on the destruction of microaggregates of carbonate chernozem [Vliyaniye ionno-solevogo sostava pochvennogo karbonata na razrusheniye mikroagregatovatnogo chernozema] Pochvovedenie. 1994; 1:5160.
45. Abbas A, Khan S, Hussain N, Hanjra MA, Akbar S. Characterizing soil salinity in irrigated agriculture using a remote sensing approach. Physics and Chemistry of the Earth. Part A/B/C/. 2013;(55-57):1-8.
46. Allbed A, Kumar L, Aldakheel YY. Assessing soil salinity using soil salinity and vegetation indices derived from IKONOS high-spatial resolution imageries: Applications in a date palm dominated region Geoderma. 2014;(230-231):1-8.
47. Bahmaei A, Albaji M, Naseri AA, Varjavand P. Effect of irrigation type and interval on soil salinity in clay soils in Ahvaz, Iran Arabian Journal of Geosciences. 2020;13:1151.
48. Chi Y, Sun J, Liu W, Wang J, Zhao M. Mapping coastal wetland soil salinity in different seasons using an improved comprehensive land surface factor system Ecological Indicators. 2019;(107):105517.
49. Copernicus Open Access Hub. 2020, available at
Monitoring soil salinity via remote sensing technology under data scarce conditions: a case study from Turkey // Ecological Indicators. Vol. 74. P. 384-391.
55. Hasanlou M., Eftekhari K. 2019. Retrieval of soil salinity from Sentinel-2 multispectral imagery // European Journal of Remote Sensing. Vol. 52:1. P. 138-154.
56. Hassania A., Azapagica A., Shokrib N. 2020. Predicting long-term dynamics of soil salinity and sodicity on a global scale // PNAS. Vol. 117. No. 52. P. 33017-33027.
57. Hick P.T., Russell W.G.R. 1990. Some spectral considerations for remote sensing of soil salinity // Australian Journal of Soil Research. Vol. 28. No. 3. P. 417-431.
58. Hihi S., Rabah Z.B., Bouaziz M., ChtourouM.Y., Bouaziz S. 2019. Prediction of Soil Salinity Using Remote Sensing Tools and Linear Regression Model // Advances in Remote Sensing. Vol. 8. P. 7788.
59. Huete A.R. 1988. Soil-Adjusted Vegetation Index (SAVI) // Remote Sensing of Environment. Vol. 25. P. 295-309.
60. IqbalF. 2011. Detection of salt affected soil in rice-wheat area using satellite image // African Journal of Agricultural Research. Vol. 6 (21). P. 4973-4982.
61. Jiang H., Shu H. 2018. Optical remote-sensing data based research on detecting soil salinity at different depth in an arid-area oasis. Xinjiang, China // Earth Science Informatics. P. 1-14.
62. Jimenez-Aguirre M.T., Isidoro D., Uson A. 2018. Soil variability in La Violada Irrigation District (Spain): II Characterizing hydrologic and salinity features // Geoderma. Vol. 311. P. 67-77 [Available at www. elsevier. com/locate/geoderma (Date of Access 16.02.2021)].
63. Kaufman Y.J., Tanre D. 1992. Atmospherically resistant vegetationin index (ARVI) for EOS-MODIS // Transactions on Geoscience and Remote Sensing. Vol. 30. P. 261-270.
64. Kauth R.J., Thomas G.S. 1976. The tasseled cap - a graphic description of the spectral-temporal development of agricultural crops
https://scihub.copernicus.eu/dhus/#/home, accessed 10/04/2021).
50. Crippen RE. Calculating the Vegetation Index Faster Remote Sensing of Environment. 1990;(34):71-73.
51. Dwivedi RS. Monitoring of salt-affected soils of the Indo-Gangetic alluvial plains using principal component analysis International Journal of Remote Sensing. 1996;(17)10:1907-1914.
52. Earth Explorer. 2020, available at https://earthexplorer.usgs.gov/, accessed 10/04/2021).
53. Forests of Randomized Trees. 2020, available at https://scikit-learn.org/stable/modules/ensemble.html#fo rest, accessed 15/04/2021).
54. Gorji T, Sertel E, Tanik A. Monitoring soil salinity via remote sensing technology under data scarce conditions: a case study from Turkey Ecological Indicators. 2017;(74):384-391.
55. Hasanlou M, Eftekhari K. Retrieval of soil salinity from Sentinel-2 multispectral imagery European Journal of Remote Sensing. 2019;(52)1:138-154.
56. Hassania A, Azapagica A, Shokrib N. Predicting long-term dynamics of soil salinity and sodicity on a global scale PNAS. 2020;(117)52:33017-33027.
57. Hick PT, Russell WGR. Some spectral considerations for remote sensing of soil salinity Australian Journal of Soil Research 1990;(28)3:417-431.
58. Hihi S, Rabah ZB, Bouaziz M, Chtourou MY, Bouaziz S. Prediction of Soil Salinity Using Remote Sensing Tools and Linear Regression Model Advances in Remote Sensing. 2019;(8):77-88.
59. Huete AR. Soil-Adjusted Vegetation Index (SAVI) Remote Sensing of Environment. 1988;(25):295-309.
60. Iqbal F. Detection of salt affected soil in rice-wheat area using satellite image African Journal of Agricultural Research. 2011;(6)21:4973-4982.
61. Jiang H, Shu H. Optical remote-sensing data based research on detecting soil salinity at different depth in an arid-area
as seen by LANDSAT // Proceedings of the Symposium on machine processing of remotely sensed data. Purdue University, West Lafayette, Ind., June 29-July 1, 1976 (A77-15051 04-43). P. 4B-41 to 4B-51.
65. Komissarov A., Safin Kh., Ishbulatov M., Khafizov A., Komissarov M. 2019. Irrigation as means to reduce the risks of agricultural productionin the South Ural // Bulgarian Journal of Agricultural Science. Vol. 25 (Suppl. 2). P. 149-157.
66.Manchanda M.L., Lyes H.S. 1983. Use of imagery and Aerial photographs for delineation and categorization of salt-affected soils of part of North-West India // Journal of the Indian Society of Soil Science. Vol. 31. No. 2. P. 263-271.
67.Masoud A.A., Koike K., Atwia M.G., El-Horiny U.U., Gemail K.S. 2019. Mapping soil salinity using spectral mixture analysis of Landsat 8 OLI images to identify factors influencing salinization in an arid region // International Journal of Applied Earth Observation and Geoinformation. Vol. 83. P. 101944.
68. Myers V.J. 1983. Remote sensing application in agriculture (salinity) // Manual of Remote Sensing. 2nd ed. USA. P. 2111-2228.
69. Myers V.J., Carter D.L., Rippert W.G. 1966. Remote sensing for estimate soil salinity // Journal of the Irrigation and Drainage Division. No. 4. P. 59-68.
70. Ren D., Wei B., Xu X., Engel B., Li G., Huang Q., Xiong Y., Huang G. 2019. Analyzing spatiotemporal characteristics of soil salinity in arid irrigated agro-ecosystems using integrated approaches // Geoderma. No. 356. P. 113935 [Available at www.elsevier.com/locate/geoderma (Date of Access 16.02.2021)].
71. Richardson J.A. 1976. Detection of soils with sky lab multispectral, scanner date program // Engineering and Remote Sensing. Vol. 5. No. 42. P. 679-684.
72. Rouse J.W., Haas R.H., Schell J.A., Deering D.W. 1973. Monitoring vegetation systems in the Great Plains with ERTS // Third ERTS Symposium, NASA SP-351 I.
oasis. Xinjiang, China Earth Science Informatics. 2018:1-14.
62. Jimenez-Aguirre MT, Isidoro D, Uson A. Soil variability in La Violada Irrigation District (Spain): II Characterizing hydrologic and salinity features Geoderma. 2018;(311):67-77, available at www. elsevier. com/locate/geoderma, accessed 16/02/2021).
63. Kaufman YJ, Tanre D. Atmospherically resistant vegetation index (ARVI) for EOS-MODIS Transactions on Geoscience and Remote Sensing. 1992;(30):261-270.
64. Kauth RJ, Thomas GS. The tasseled cap -a graphic description of the spectral-temporal development of agricultural crops as seen by LANDSAT. Proc. of the Symposium on machine processing of remotely sensed data, June 29-July 1, 1976, Purdue University, West Lafayette, Ind. 1976:4B-41-4B-51.
65. Komissarov A, Safin Kh, Ishbulatov M, Khafizov A, Komissarov M. Irrigation as means to reduce the risks of agricultural productionin the South Ural Bulgarian Journal of Agricultural Science. 2019;(25)2:149-157.
66. Manchanda ML, Lyes HS. Use of imagery and Aerial photographs for delineation and categorization of salt-affected soils of part of North-West India Journal of the Indian Society of Soil Science. 1983;(31)2:263-271.
67. Masoud AA, Koike K, Atwia MG, El-Horiny MM, Gemail KS. Mapping soil salinity using spectral mixture analysis of Landsat 8 OLI images to identify factors influencing salinization in an arid region International Journal of Applied Earth Observation and Geoinformation. 2019;(83):101944.
68. Myers VJ. Remote sensing application in agriculture (salinity) Manual of Remote Sensing. 2nd ed. USA, 1983:2111-2228.
69. Myers VJ, Carter DL, Rippert WG. Remote sensing for estimate soil salinity Journal of the Irrigation and Drainage Division. 1966;4:59-68.
70. Ren D, Wei B, Xu X, Engel B, Li G, Huang Q, Xiong Y, Huang G. Analyzing
P. 309-317.
73. Shahid S.A., Abdelfattah M.A., Taha F.K. 2012. Developments in Soil Salinity Assessment and Reclamation. Springer. 801 p.
74. Sinanuwong S., Wichaidisdha P., Trakuldist P. 1980. The use of Landsat imagery for soil salinity study in the N-E of Thailand // Thai Journal of Agricultural Science. Vol. 13. No. 3. P. 227-237.
75. Singh A.N. 1997. Monitoring change in the extent of salt-affected soils in northern India // International Journal of Remote Sensing. Vol. 15. No. 16. P. 3173-3187.
76. Szatmari G., Bakacsi Z., Laborczi A., Petrik O., Pataki R., Tdth T., Pasztor L. 2020. Elaborating Hungarian Segment of the Global Map of Salt-Affected Soils (GSSmap): National Contribution to an International Initiative // Remote Sensing. No. 12. P. 4073.
77. Wang J., Ding J., Yu D., Ma X., Zhang Z., Ge X., Teng D., Li X., Liang J., Lizaga I., Chen X., Yuan L., Guo Y. 2019. Capability of Sentinel-2 MSI data for monitoring and mapping of soil salinity in dry and wet seasons in the Ebinur Lake region, Xinjiang, China // Geoderma. Vol. 353. P. 172-187.
78. Zaman M., Shahid S.A., Heng L. 2018. Guideline for Salinity Assessment, Mitigation and Adaptation Using Nuclear and Related Techniques. International Atomic Energy Agency, Springer Nature Switzerland AG. 164 p.
spatiotemporal characteristics of soil salinity in arid irrigated agro-ecosystems using integrated approaches Geoderma. 2019;356:113935, available at www. elsevier. com/locate/geoderma, accessed 16/02/2021).
71. Richardson JA. Detection of soils with sky lab multispectral, scanner date program Engineering and Remote Sensing. 1976;(5)42:679-684.
72. Rouse JW, Haas RH, Schell JA, Deering DW. Monitoring vegetation systems in the Great Plains with ERTS. Third ERTS Symposium, NASA SP-351 I, 1973:309-317.
73. Shahid SA, Abdelfattah MA, Taha FK. Developments in Soil Salinity Assessment and Reclamation. Springer, 2012:801.
74. Sinanuwong S, Wichaidisdha P, Trakuldist P. The use of Landsat imagery for soil salinity study in the N-E of Thailand Thai Journal of Agricultural Science. 1980;(13)3:227-237.
75. Singh AN. Monitoring change in the extent of salt-affected soils in northern India International Journal of Remote Sensing. 1997:(15)16:3173-3187.
76. Szatmari G, Bakacsi Z, Laborczi A, Petrik O, Pataki R, Toth T, Pasztor L. Elaborating Hungarian Segment of the Global Map of Salt-Affected Soils (GSSmap): National Contribution to an International Initiative Remote Sensing. 2020;12:4073.
77. Wang J, Ding J, YuD, Ma X, Zhang Z, Ge X, Teng D, Li X, Liang J, Lizaga I, Chen X, Yuan L, Guo Y. Capability of Sentinel-2 MSI data for monitoring and mapping of soil salinity in dry and wet seasons in the Ebinur Lake region, Xinjiang, China Geoderma. 2019;(353):172-187.
78. Zaman M, Shahid SA, Heng L. Guideline for Salinity Assessment, Mitigation and Adaptation Using Nuclear and Related Techniques. International Atomic Energy Agency, Springer Nature, Switzerland AG, 2018:164.
