Using GIS technologies to determine the weediness of agricultural crops in the example of the Akmola region

. Data from the PlanetScope remote sensing mission were used to determine the infestation of crops in the Akmola region. Data were obtained from 17 districts and two municipal cities. The statistical analysis of objects is shown, and the minimum, maximum, average, and standard deviation of the indicators of the studied sown areas in the Akmola region are revealed. Spring-summer and summer-autumn surveys of weed infestation of agricultural crops showed that despite the implementation of regular agrotechnical and chemical protective measures, the general condition of crops was determined as satisfactory (medium). The study shows that a complete database of information about agricultural, arable, and fallow fields could be easily formed with GIS technologies, and survey data of digitized crops in the Akmola region are available in real time for most users.


Introduction
According to the Bureau of National Statistics of the Agency for Strategic Planning and Reforms of the Republic of Kazakhstan, the gross harvest of grains and legumes for 2018-2020 in the Republic of Kazakhstan amounted to 2018 -20,273.7 thousand tons, 2019 -17,428.6 thousand tons, 2020 -20,065.3thousand tons [2].
Statistical data analysis states that wheat crops occupy a significant part of the harvest.In 2020, three regions became leaders in wheat-producing: Akmola (4,127.6 thousand tons), Kostanay (3,455.0thousand tons), and North Kazakhstan (3,299.8thousand tons) [2].
The preservation of high-quality and safe raw materials is one of modern Kazakhstan's most important strategic tasks.Due to the improvement of agricultural technologies over the past decade, there has been a positive trend in wheat yields in most areas of Akmola and other regions.
It should be noted that the Turkestan, Zhambyl, and Almaty regions showed the best results in terms of wheat yield in 2021, 19.9, 19.8, and 19.7 centners per hectare, respectively (Table 1).Obtaining data on agriculture, mainly field weeds, yields, and soil conditions, is timeconsuming and costly, so proper statistical analysis of the data is often a challenge for producers [5].However, many studies state that obtaining data using remote sensing and satellite imagery may significantly change the game, this imagery becomes an alternative data source for determining crop productivity indicators.They are easy and relatively inexpensive to obtain, and they are statistically reliable compared with ground data [6,7,8,9].
Satellite-based remote sensing is a high-performance tool for obtaining agricultural information with spatial and temporal resolution [10].
Crop remote sensing provides mapping, monitoring, and modeling of crop types, epidemics, and crop quality and quantity.The resulting maps of individual agricultural fields are considered informative [11].
In 2015, two optical spacecraft, KazEOSat-1 (high resolution -1 m) and KazEOSat-2 (medium-resolution -6.5 m), were put into operation in Kazakhstan, with which a space remote control system was created and successfully operated remote sensing.The system is designed to conduct space surveys of the Earth's surface and provide operational monitoring information to solve different problems in the economic sector of the Republic of Kazakhstan (Figure 1) [1].
According to statistical data from the official source [3], developed as part of the resolutions from the Kazakhstan President, K.K. Tokayev on the introduction of digital monitoring and control over the rational use of agricultural land, the share of digitized land, the total area of fields in the Akmola region amounted to 5,957,097 hectares, the area of digitized land -6,136,397.89ha, the number of land users -3,556, the total number of electronic land -27 418.In the context of districts, the share of digitized lands from the entire area is 96.7 -100% (Table 2) [3].

Fig. 1. Scheme of satellite estimation of sown areas of crops.
Using the data obtained from the PlanetScope geoinformation system, it is possible to get high-quality maps since the system has a high spatial resolution (3 m) [12].Compared to the traditional data collection approach, the advantage of remote sensing in agricultural management is undeniable [13].In addition, PlanetScope satellites are relatively inexpensive to manufacture, and available, which allows you to create a collection of images with both high spatial and temporal resolution (up to 1 week) at a lower cost [14].
In general, previous studies highlight the effectiveness of using space-based and ground-based monitoring data.The main purpose of this article is to explore the statistical indicators of the collected data using geographic information systems in determining the weediness of crops in the Akmola region.

Materials and methods
This study focuses on agricultural crops in the Akmola region.The Akmola region includes 17 districts and 2 municipalities.For further determination of weediness, the fields are measured by regular intervals along the largest diagonal of frames that are divided into sections of 25x100 cm.Inside the structure, the number of weeds by species is counted.Depending on the degree of infestation (number of weeds per square meter), the surveyed areas are grouped according to the following gradations: 1-5; 6-15; 16-50; 51-100; more than 100.After the registration of weeds and the determination of their biological groups, mapping the weediness of the fields begins [4].
The first stage is a spring-summer inspection of crops in the phase of weed regrowth for short-term forecasting of weediness and chemical weeding.The second stage -summerautumn surveys before harvesting, is for long-term forecasting and planning of agrotechnical and chemical protective measures.
When assessing agricultural areas on the territory of Kazakhstan, space images from the following sources are used: images from the KazEOSat-2 remote sensing satellites (with a spatial resolution of 6.5 m/pixel), PlanetScope (with a spatial resolution of 3.5 m/pixel), images from the remote sensing satellite Landsat (with a spatial resolution of 30 m/pixel), images from the remote sensing satellite Sentinel-2 (with a spatial resolution of 10 m/pixel) Figure 2 shows a digital map of agricultural land in the Akmola region, which includes 6 zones: fallow (orange), not plowed (red), lying fallow (blue), arable land (sown) (yellow), arable land (unsown) (dark green) and hayfield (light green).To obtain reliable information on the weed infestation of agricultural land, the authors have collected data from the professionals of the "Republican Methodological Center for Phytosanitary Diagnostics and Forecasts" of the State Inspection Committee in the Agroindustry Complex of the Ministry of Agriculture of the Akmola Region [15].The collected sample data on the weediness of agricultural land at different stages of ripeness from 11 districts of the Akmola region are shown in Table 3.

Results and discussion
The analysis of the field survey data showed in Table 3 that the wheat crops of the Akkol, Astrakhan, Bulandinsky, Zerenda, Sandyktau, and Atbasar districts in the stage of milky ripeness were mostly infested to an average degree, and the visual yield for wheat was 8.8-15.1 centner/ha, the condition of the crops is good, while the barley crops mainly were infested to a low degree, the yield of barley was 13.2-19.8centner/ha, and the condition of the crops is good.
The main methods of weed control are agrotechnical and chemical.When using mechanical tillage, using tools with paw working parts is necessary.With a delay in the emergence of weeds, intermediate mechanical tillage before sowing is effective.It is

Conclusions
In 2020, three regions became leaders in wheat-producing: Akmola (4,127.6 thousand tons), Kostanay (3,455.0thousand tons), and North Kazakhstan (3,299.8thousand tons).On the other hand, in terms of wheat yield characteristics, Turkestan, Zhambyl, and Almaty regions were in the lead in 2021, with 19.9, 19.8, and 19.7 centners per hectare, respectively.Data collection is a rather laborious process, that could be significantly facilitated if the digitalization of land is present.According to statistics, digitized land in the Akmola region ranged from 96.7 to 100%.The digital map of agricultural land in the Akmola region includes 6 zones: fallow (orange), not plowed (red), lying fallow (blue), arable land (sown) (yellow), arable land (unsown) (dark green) and hayfield (light green).At the milky ripeness stage, the condition of wheat crops in the Akkol, Astrakhan, Bulandinsky, Zerenda, Sandyktau, and Atbasar regions was mainly good, with the moderate weed infestation and the visual yield for wheat varying from 8.8 to 15.1 c/ha.The condition of barley crops in the same districts also showed good results: weed infestation is weak, and barley yield ranges from 13.2 to 19.8 q/ha.The statistics of filtered data on the state of agricultural lands in the Akmola region showed that in Zharkainsky, Atbasarsky, Akkolsky, and Zerenda districts, the weeds accounting and data introduction was the most systematic and frequent, which in turn makes it possible to compile a detailed map of their infestation.
Development of digital maps of agricultural land in all regions of the Republic of Kazakhstan based on remote sensing data of medium resolution Detection of sown fields with the allocation to the category of the subclass "sowing" of the class "arable land" of the territory of the region (search, filtering, and primary processing of remote sensing data, downloading space images, mapping agricultural land, creating field masks, classification by land categories) Verification of data from space monitoring of the sowing campaign with groundbased observations (drawing up a plan-scheme of the route of field surveys, field survey trips, office processing) Determination of sowing dates for each identified sowing area based on the use of current remote sensing data of the region (calculation of vegetation indices, visual interpretation of space images) Development of maps-schemes of sown areas of crops in all regions of the Republic of Kazakhstan A summary of sown areas of crops in the context of districts of the region (calculation of areas, creation of thematic maps, publication of data in a geoservice) [1].The images were taken in 2021.A controlled method was used to separate image pixels into different classes in the work.Supervised classification combines pixels into categories based on user data for training.The training data can come from an imported ROI file or user-created ROIs in the image.After the satellite image is loaded, it is processed for further use for classification.Next, noise is eliminated, and atmospheric correction is performed in order to reduce the influence of natural factors on the spectral data [1].

Table 2 .
Digitization of arable land by districts of the Akmola region as of 2021 [3].