An Estimation Model of Emissions from Burning Areas Based on the Tier Method

Abstract

The emissions of particulates from burning agricultural fields threaten the environment and human health, contributing to air pollution and increasing the risk of respiratory and cardiovascular diseases. An analysis of total suspended particulate (TSP), PM2.5, and PM10 emissions from crop residue burning is presented in this study. A primary goal is to improve emission estimation accuracy by integrating satellite imagery from modes of Moderate Resolution Imaging Spectroradiometers (MODIS) and Visible Infrared Imaging Radiometers (VIIRS) with traditional data. Particulate emissions were estimated using Tier 1 and Tier 2 methodologies outlined in the EEA/EMEP Emission Inventory Guidebook based on thermal anomaly data from satellite observations. According to the findings, burning wheat, maize, barley, and rice residue accounts for most emissions, with significant variations identified in India, China, and the United States. The variations highlight the need for a location-specific approach to emission management. Particulate emissions cause adverse environmental and health impacts, which can be minimized by targeting mitigation strategies at key emission hotspots. The research provides important insights to inform policymakers and support developing strategies to reduce fine particulate agricultural emissions.

1. Introduction

1.1. Air Quality Standards for Suspended Particulate Matter

In agriculture and industry, fine particles of matter, called particulates, are distributed in the air [1]. Particulates are usually produced in agricultural operations and processes from controlled and uncontrolled sources. Airborne particulates harm living organisms, including people [2]. As a result, the respiratory and cardiovascular systems are exposed to various threats [3]. Particle pollution is defined as the chemical or biological alteration of solids, liquids, or gases in the air at a level that is harmful to people, animals, plants, and property and is a nuisance [2]. Regulations on air quality and cleanliness in the European Union are governed by Directive 2008/50/EC [4]. This directive outlines how to assess corrective actions and targets to establish target and limit values for PM2.5 concentrations in Europe. A separate indicator is specified for urban areas, requiring public notification of dust hazards [5,6].

Poland is implementing legal changes to improve the country’s air quality. A single reading does not determine a standard exceedance but rather statistically over 24 h. PM2.5 dust measurements are complex [1]. The USA’s NAAQS (National Ambient Air Quality Standards) describes determining threshold concentrations for some pollutants. The NAAQS determined PM2.5 concentrations using two measurement procedures [7]. The first type of measurement uses air pollen saturation over 24 h, with the average value not exceeding 35 µg/m³. The second type of measurement is made annually, where the annual average concentration must not exceed 15 µg/m³ [8]. Dust concentrations above a threshold are considered at-risk areas (NAAQS) where higher concentrations may occur [9]. Several parts of California, Arizona, and southern Texas are considered NAAQS at-risk areas due to high PM2.5 dust levels. Monitoring was conducted in areas belonging to large farms in the California San Joaquin Valley and south Arizona [10].

1.2. Estimatimating the PM10 and PM2.5 Pollution from the Ground

Agricultural activity interferes with the natural environment. The emission of particulate matter (PM) during the burning of agricultural fields is a critical environmental threat. PM emissions from field burning can be measured with specialized tools that categorize particles by size. Fires in fields can also occur as unplanned events during droughts. The causes of fires include high temperatures, low humidity, and strong winds [11]. Burning grasses and crop residues to prepare soil for new crops also results in some fires. It may improve soil properties in the short term by releasing minerals, but such actions are hazardous [12]. Uncontrolled fires can spread over vast areas. Burning grass also damages the environment, destroying biodiversity, degrading soil, and releasing large amounts of greenhouse gases [13]. The practice of burning fields is prohibited in many countries, including Poland. However, it is still practiced due to tradition and a lack of awareness of its consequences [14].

Air pollution indicators (EF_pollutant) are commonly used to assess particulate matter emissions. Preliminary separators provide quick and precise results since they separate particulates by their aerodynamic properties. The proportions of PM10 and PM2.5 can be calculated as a function of total PM emissions. Monitoring particulate matter (PM) emissions from agricultural burning relies on a combination of precise, real-time active sensors and simpler passive devices that are wind-dependent [15]. This monitoring is governed by Polish laws and Directive 2008/50/EC [4]. Optical sensors like Datar and Grimm examine the scattering or absorption of light to measure PM. Gravimetric measurements are precise reference standards. TEOM devices and particle analyzers are widely used for monitoring PM2.5 and PM10. Passive sensors like MWAC, BSNE, and SUSTRA collect particles using wind. MWAC has 90% or more accuracy for fine particles, and BSNE performs better at higher wind speeds [16,17]. These methods allow for assessing the environmental impact of agricultural burning [17].

1.3. Estimatimating the TSP, PM10 and PM2.5 Pollution from Remote Sensors

Using satellite imagery to estimate TSP, PM10, and PM2.5 pollution allows for precise analysis of particulate matter emissions from biomass burning, including field burning, which significantly impacts air quality [18]. Burning can cause dust and smoke emissions in dry and windy conditions. During the California fires, dry grasses and strong winds contributed to rapid fire spread. Satellite images, such as those from Landsat 9 and MODIS, can identify fire-affected areas, estimate their size, and measure their spatial distribution [19,20,21,22]. Similar phenomena are observed in other regions where biomass burning, both controlled and uncontrolled, contributes to particulate matter emissions and their long-distance movement. Data allows for assessment of the effect of emissions on human health and the environment and supports efforts to reduce emissions [23]. Satellite imagery facilitates monitoring active fires and forecasting vulnerable areas, which helps us to prevent fires and plan strategic environmental policies [24,25,26].

Integrating satellite data and dust emission models helps us to understand the dynamics of biomass burning processes, reducing their adverse effects. This paper uses satellite imagery to examine the spatial variation in particulate emissions from field burning. The methodology employs specific coefficients to estimate particulate emission levels, such as PM2.5 and PM10, enabling detailed and quantifiable analysis. The findings offer benefits, including identifying primary sources of particulate pollution and providing data to inform emission control strategies. Particulate matter harms human health, and this analysis supports mitigation efforts. The results will help policymakers and stakeholders identify effective measures for reducing emissions and optimizing agricultural practices [27]. This integrated approach combines technological advancements and spatial analysis to address particulate pollution challenges.

1.4. Emission Estimation Models

The fire emissions in this study were estimated using Tier 1 and Tier 2 estimation methods, commonly used in emissions inventories. According to the Intergovernmental Panel on Climate Change (IPCC) guidelines, these methods are used in environmental analysis because of their transparency and applicability under various conditions, including limited input data availability [12]. Tier 1 uses global emission factors and averages of combustion parameters for different ecosystems to simplify emissions estimates. Tier 2 uses more detailed data specific to a region or biomass type, resulting in more precise estimates. The methods were selected based on data availability and their wide application in research on fire emissions.

Other approaches are available to estimate fire emissions from satellite data. The Fire Radiative Power (FRP) method uses satellite sensors like MODIS and VIIRS to document the thermal energy emitted by fires. FRP directly links fire intensity to emission rate, offering an alternative to methods based on burned areas [28]. Van Der Werf et al., (2017) incorporated burned area data, active fires, and biogeochemical models into the Global Fire Emissions Database (GFED) to estimate emissions [29]. Global fire emissions can be visualized, but a lack of detailed combustion data may hinder sophisticated modeling.

Atmospheric chemistry models combine satellite data on fire activity with meteorological and atmospheric transport models to determine the impact of emissions on air quality and their long-range transport [30]. Inverse modeling uses satellite data on CO₂, CO, and NO_x concentrations to estimate fire emissions, enabling independent validation of emission inventory results [31]. Although these methods were not used in this study, they can be a helpful alternative or complementary technique for estimating emissions. Tiers 1 and 2 were chosen for their simplicity and data availability, making their emissions analysis efficient in a resource-limited environment. Including alternative methods in the future could enhance the accuracy of estimates and allow for additional validation, contributing to a more accurate model of fire’s environmental impact.

The study aims to improve the accuracy of particulate emission estimates from agricultural field burning rather than forest fires. It analyzes total suspended particulate (TSP), PM2.5, and PM10 emissions by integrating satellite imagery from MODIS and VIIRS. Based on satellite measurements, the researchers apply Tier 1 and Tier 2 methodologies outlined in the Emission Inventory Guidebook by the EEA and EMEP. In previous studies, Tier 1 methods have often been used to estimate emissions, which has led to either under- or overestimation. An important novel feature of this manuscript is the combination of satellite-derived burned area data with crop-specific parameters under Tier 2 methodology, enabling a more detailed and regionally adjusted estimation of emissions. Agricultural burning can be better controlled and mitigated by utilizing this approach, which helps to better quantify emission sources.

2. Materials and Methods

2.1. Scanning Satellite Image

Satellites equipped with advanced thermal sensors play an essential role in detecting thermal anomalies on the surface of the Earth. Detecting heat sources from cosmic space is made possible by systems such as MODIS [19,20,21,22] and VIIRS (Visible Infrared Imaging Radiometer Suite) [25,26] which provide valuable information for agriculture, fire monitoring, and environmental studies. Thermal satellite sensors can identify thermal anomalies on satellite images as red squares. As a result of fire, hot smoke, agricultural activities, or other processes that release heat, these anomalies can occur. It is important to note that the size and resolution of the detected areas vary depending on the satellite system used. Fire products MCD14ML and VNP14IMG were used by MODIS and VIIRS [26]. Based on MODIS confidence levels of 70% and VIIRS fire radiant power (FRP) thresholds of 5 MW, fire pixels were detected based on thermal anomalies. Fire detection is exact, but not every detection indicates a fire has spread throughout the area.

The difference in spatial resolution between MODIS (1 km) and VIIRS (375 m) impacts fire detection capabilities and burnt area mapping accuracy. Due to its lower resolution, MODIS can detect large fires more effectively, but smaller or fragmented fires may be overlooked. As a result, VIIRS can provide a much more detailed image of smaller fires than MODIS, which could allow it to identify fires that MODIS might otherwise overlook. Emission models must account for these discrepancies when integrating both datasets, as they affect the calculation of burned areas and total emissions.

An essential element of estimating particulate emissions of PM10 and PM2.5 is determining the land area burned by fires from the data obtained. To estimate emissions for this purpose, a Tier 1 method was used based on calculating the burned area and applying appropriate emission factors [32]. The method described in the following subsections allows for a quantitative assessment of emissions based on the spatial extent of fires. Information from satellites can be used to assess fire emissions’ environmental and health effects. The acquired images enable monitoring regions with a high fire risk and identifying areas needing special attention in terms of ecological management. Through integrating satellite technology with Tier 1 methodologies, precise estimates of the impact of fires on air quality and strategies for reducing emissions can be made. In this way, the provided analysis can be a powerful tool for analyzing the environmental impact of wildfires [33,34]. The schematic of the workflow for estimating emissions from burned areas was presented in Figure 1.

The satellite images used in this study were obtained from publicly available NASA databases, specifically from MODIS and VIIRS. These sensors are widely used for fire detection because they capture thermal anomalies associated with active fires. MODIS provides global fire observations twice daily at a 1 km × 1 km resolution [21], while VIIRS offers higher spatial precision with a 375 m × 375 m resolution [22], allowing for more detailed fire mapping. Thermal satellite sensors can identify thermal anomalies on satellite images as red squares, which can result from fires, hot smoke, agricultural activities, or other heat-releasing processes. However, it is important to note that not every detection necessarily indicates fire spread over an entire area.

The first Californian case was a forest fire, as confirmed by MODIS and VIIRS satellite imagery. It corresponds to the Carr Fire (July–August 2018), which burned over 92,936 hectares in Northern California. As compared to crop residue burning, which primarily emits particulate matter (PM2.5 and PM10), wildfires in California produce higher levels of aromatic hydrocarbons and nitrogen oxides during the dry season. Based on satellite-based estimates, PM2.5 emissions from this event were 3.474 tons per 100,000 hectares, which is within expectations for forest fires of this size. Interpreting results requires recognizing this distinction since the impact of emissions on air quality and public health varies significantly depending on the kind of fire and its combustion characteristics.

To estimate particulate emissions of PM10 and PM2.5, it is essential to determine the land area burned using the obtained satellite data. For this purpose, a Tier 1 method was applied, which involves calculating the burned area and applying appropriate emission factors. The methodology described in the following subsections enables a quantitative assessment of emissions based on the spatial extent of fires. Satellite data are valuable in evaluating fire emissions’ environmental and health impacts. The acquired imagery allows for monitoring fire-prone regions and identifying areas that require special attention in terms of ecological management. By integrating satellite technology with the Tier 1 methodology, precise estimations of fire impacts on air quality and strategies for reducing emissions can be developed. This combination of satellite-based fire monitoring and emission estimation methods makes the presented analysis a powerful tool for studying the environmental impact of agricultural residue burning and wildfires [25,26].

2.2. Picture Bit-Depth Conversion

A custom-designed computer program converted a source image into its final form. The use of a specially designed program can streamline and improve the process of image processing. The method provides images to be converted into 16-bit files, permanently altering their color [35]. The bit-depth conversion was performed using a Delphi program, stretching a 16-bit histogram into an 8-bit histogram. Thermal signals relevant to detecting burned areas were extracted using red-channel segmentation [36]. Simplification offers significant advantages, such as clearly defining distinct fields within an image. The program makes subsequent computations more efficient by reducing image complexity and facilitating precise segmentation and analysis. Based on predefined color characteristics, the program analyzes the image efficiently, identifying and isolating regions of interest. Applications requiring image regions to be differentiated from one another according to specific properties, such as brightness or color distribution, typically use this functionality. Through the use of a dedicated filter and specially adapted tools, the image is reduced to bits [13,37]. A custom algorithm quantifies and categorizes red color intensity before being converted to pixel colors. The pre-processing step identifies areas of interest early in the workflow, enhancing accuracy. Delphi programming language was used to develop the program’s interface and make it user-friendly. The algorithm for converting the bit depth of a screen is presented in Figure 2.

The bit-depth conversion process significantly reduces color saturation information within an image. This process removes hue and saturation data, leaving only luminance data representing each pixel’s intensity. By simplifying their format to 16 bits, images become more efficient to analyze computationally and reduce complexity [35,36]. Based on a custom algorithm embedded in the program, pixel intensities are analyzed and classified based on the reduced bit depth, enabling the program to segment image regions accurately. It is still possible to reliably analyze luminance data within the scale’s limits despite the lack of accuracy introduced by simplification. Using the combined data points, we can estimate proportions within an image, like the pixel intensity distribution. The methodology improves computational efficiency and accuracy, especially in applications that require detailed segmentation, area evaluation, or pattern recognition. Using this approach is particularly beneficial when describing the characteristics of specific regions for analysis is necessary. The methodology is a valuable tool for analyzing and segmenting images accurately.

Bit-depth conversion and point-to-point scaling are necessary to ensure computation efficiency during image processing. Distortions caused by these transformations can affect the accuracy of burned area estimations, thus affecting emission calculations. The loss of fine details that are critical to identifying burned regions could result from bit-depth conversion, which reduces the range of pixel values. Through point-to-point scaling, spatial resolution is adjusted through interpolation, resulting in changes to fire perimeters and spatial distortions. The impact of these errors on emission estimates was assessed by conducting a sensitivity analysis using high-resolution satellite imagery. Different bit depths and scaling conditions were used to compare burned area detection results before and after processing. Analysis was conducted to test multiple reduction strategies (e.g., 16-bit to 8-bit conversion) and scaling methods (e.g., bilinear, bicubic, and nearest neighbor). Devices observed in burned area estimations were propagated through the emission calculation framework to quantify their impact on total emissions. By utilizing satellite-derived fire products, repeated test cases quantified these sources of uncertainty. The results of sensitivity analysis for the burned area and emission estimations using variations in image processing, such as bit-depth conversion and point-to-point spatial scaling, are presented in

Table 1. Potential sources of error in image processing and their impact on emission estimates.

Error Source	Description	Estimated Impact on Burned Area, %	Estimated Impact on Emission Estimates, %
Bit-depth conversion	Loss of fine details affecting fire area segmentation	4.2	4.8
Point-to-Point Scaling	Interpolation-induced spatial distortions	3.8	4.2
Combined Impact	Cumulative effect of both processes	7.5	8.2

.

Bit-depth conversion can result in a deviation of 4.2% in the burned area, while point-to-point scaling can result in an approximately 3.8% variation in the burned area. The cumulative error in both processes can result in a deviation from the emission estimate of up to 8.2%, depending upon fire characteristics and landscape conditions. To illustrate how image processing affects burned area estimations, we use a bit-depth conversion dataset and a point-to-point scaling example. The final corrected estimate, obtained through validation and calibration methods, illustrates the reduction in error brought about by the processing procedure. An example of burned area estimation variability due to image processing errors is presented in

Table 2. Example of burned area estimation variability due to image processing errors.

Method	Burned Area, ha	Error, %
Raw Satellite Data	1000	0
Bit-Depth Conversion	950	−5
Point-to-Point Scaling	970	−3
Final Corrected Estimate	990	−1

.

Bit-depth conversion and point-to-point scaling can reduce burned areas by as much as 5% and 3%, respectively. The corrected estimate brings the deviation down to 1% after validation with independent datasets and ground-based fire observations, highlighting the effectiveness of the correction strategies. Several correction strategies were employed to mitigate the errors. Satellite imagery was validated with independent high-resolution datasets using a multi-resolution approach. The burned area estimates were further refined based on ground-based fire observations. Adaptive resampling methods are also useful for minimizing spatial distortions caused by scaling, and machine learning-based classification techniques can be incorporated for further improvements. Enhancing the robustness of emission estimates through optimization of these correction methods and integrating uncertainty quantification techniques should be the focus of future research.

2.3. Emission Estimation Methods

2.3.1. Tier 1 Method

Burning grasses and agricultural residues produce dust with PM2.5 and PM10 fractions, significantly affecting air quality, human health, and the environment. A Tier 1 method is recommended for estimating dust emissions from grass burning (NFR category 3.F—areas burning agricultural residues) [33,34]. It is a mathematical method that uses simplified, standardized assumptions for estimation. IPCC (Intergovernmental Panel on Climate Change) guidelines define Tier 1 as a basic level of accuracy, distinguished by its simplicity, making it particularly useful in situations without detailed input data [34]. Standard emission factors (EFs) and estimates of burned agricultural residues (activity data) are some of the key elements of the method. The method permits a quick and relatively accurate estimation of dust emissions, making it a valuable tool for monitoring and planning emission reduction activities. Calculating dust emissions involves using Equation (1).

E_pollutant = AR_{residue_burnt} · EF_pollutant

(1)

where

• E_pollutant—emission (E) of pollutants, kg/year;
• AR_{residue_burnt}—mass of burnt residues (dry matter), kg/year;
• EF_pollutant—emission factor, kg/kg s.m.

Using the formula presented here, it is possible to estimate dust emissions from field burning, which is essential for understanding how these practices impact air quality, human health, and the environment. The annual dimension of burned agricultural residues in the country must be considered when estimating dust emissions associated with burning. The activity index (AR_{residue_burnt}) incorporates data on the area of burned fields, the weight of residue per unit area, and the burn rate, which reflects how much residue is burned. Incorporating these data allows a more precise analysis of dust emissions, which is critical to developing protective strategies. The AR_{residue_burnt} activity index can be determined using Equation (2) [38].

AR_{residue_burnt} = A · M_b · C_f

(2)

where

• A—burned area, ha/year;
• M_b—mass of burnt residues (tones/ha);
• C_f—combustion coefficient.

Equation (2) provides an estimate of pollutant emissions based on the mass of burned residues (AR_{residue_burnt}). Satellite imagery from MODIS and VIIRS detected thermal anomalies associated with active fires, which led to the determination of the burned area (A). The classification of these anomalies enabled quantification of burned areas in hectares. Based on biomass yield coefficients, the mass of burned residues (M_b) was estimated. Based on agricultural statistics and previous studies on biomass availability after harvest, these coefficients are derived. The burning coefficient (C_f) represents the proportion of biomass that actually burns during a fire. Default values were used for general agricultural burning, while crop-specific values were applied in accordance with regional burning practices and literature recommendations. Calculations such as these ensure emissions are estimated based on realistic assumptions.

The Tier 1 method uses default emission factors (EFs) to calculate dust emissions from the surface burning of agricultural residues [33]. A scientific research-based approach was used to develop these guidelines. In cases where detailed local data is unavailable, these factors can be used to estimate emissions rapidly and uniformly. As wheat cultivation is prevalent, the Tier 1 method mainly developed emission factors for the combustion of wheat residues, such as straw. Based on averaged EF values for PM10, PM2.5, and total particulate matter (TSP), the values may be used as a benchmark under various climatic and agronomic conditions. The values of these emission factors and their 95% confidence intervals are provided in

Table 3. Default dust emission factors in the Tier 1 method. The emission factors (EFs) for different pollutants, including total suspended particulates (TSPs), PM10, and PM2.5, are derived from combustion events and calculated per unit of burnt residue. The confidence interval (95%) provides an indication of the variation in emission estimates.

Pollution	Emission Factor (EFpollutant), kg/kg s.m.	Confidence Interval (95%)
		Lower	Upper
TSP	0.0058	0.0045	0.0071
PM10	0.0057	0.0044	0.0071
PM2.5	0.0054	0.0042	0.0067

.

The method can be improved by prior declaration of the dry matter value based on the area of the measured field. It would be appropriate to define analyzed areas (e.g., significant crops by climate zones) based on individual countries, considering more accurate annual compilations. A conceptual demonstration of the Tier 1 method used general emission factors and large-scale fire data to estimate California’s fire emissions. The study aimed to illustrate the feasibility of including satellite-derived fire detection in emission estimations. As an alternative, the Tier 2 method was applied to other case studies that had more detailed, region-specific data on crop residue types, combustion efficiency, and local conditions. In order to improve the accuracy of Tier 2, it must be able to incorporate localized parameters, so its application to California would require a specific understanding of the burning characteristics of biomass, which was not the primary focus of this study. The comparison of the two methods in the same area could be a valuable direction for future research, particularly for improving spatial emissions modeling in wildfire-prone areas.

2.3.2. Tier 2 Method

As an extension of the Tier 1 method, the Tier 2 method provides a more accurate estimate of dust emissions because it considers additional emission factor (EF) parameters [33,34]. The Tier 2 method considers detailed local data, such as the type of agricultural residue burned, its moisture content, its dry matter content, and combustion conditions, unlike Tier 1, which assumes average values. This means it can estimate PM10 and PM2.5 emissions more accurately, which is crucial in regions where agriculture and climate are scarce. As well as allowing for better emissions mapping across locations, the method is also helpful in creating detailed dust emission inventories and supporting local environmental protection initiatives. The default dust emission factors for agricultural residue burning (NFR 3.F—surface burning with agrarian residues), according to the type of residue burned, are presented in

Table 4. Default dust emission factors used in the Tier 2 method. This table presents refined emission factors specific to various crop residues such as wheat, barley, maize, and rice. The values indicate the estimated emissions for each pollutant category (TSP, PM10, and PM2.5) and highlight variations due to different burning conditions and biomass types.

Type of Crop Residues	Crop Residues	Emission Factor (EFpollutant), kg/kg s.m.	Confidence Interval (95%)
			Lower	Upper
Wheat	TSP	0.0058	0.0045	0.0071
	PM10	0.0057	0.0044	0.0071
	PM2.5	0.0054	0.0042	0.0067
Barley	TSP	0.0078	0.0067	0.0088
	PM10	0.0077	0.0067	0.0087
	PM2.5	0.0074	0.0064	0.0085
Maize	TSP	0.006	0.0048	0.0078
	PM10	0.0062	0.0047	0.0077
	PM2.5	0.006	0.0045	0.0074
Rice	TSP	0.0058	0.0035	0.0078
	PM10	0.0058	0.0035	0.0077
	PM2.5	0.0055	0.0031	0.0074

[39].

The above table presents emission factors for wheat, barley, corn, and rice residues. It shows that each crop emits three types of pollutants, namely TSP, PM10, and PM2.5. Additionally, each emission factor has a confidence interval indicating its range of values. The table demonstrates that different types of crop residue have different emission levels for each pollutant. Generally, rice has the widest confidence interval among all contaminants, whereas barley is the most likely to generate TSP. As a result of these values, further analysis will be carried out in this research.

2.3.3. Tier 3 Method

The Tier 3 method, the most advanced approach to estimating pollutant emissions, uses detailed measurement data and source-specific modeling. The technique is not commonly used to determine particulate emissions from surface burning of agricultural residues (NFR 3.F). This is mainly due to insufficient empirical data and models that can accurately calculate dust emissions. The Tier 3 method is often used to estimate emissions of gases such as CO, NO_x, NMVOC, and SO₂, which have more detailed data and emission models available. Because particulate formation processes are complex and there is not enough data, Tier 3 is rarely used for estimating particulate emission estimates, and Tier 1 and Tier 2 methods are used instead [33,34,35]. Therefore, Tier 1 and Tier 2 estimation methods are the most common for estimating particulates.

Tier 3 methodology is more advanced and provides a more accurate approach to emission estimation. The study did not utilize it due to several practical concerns. In this research, highly detailed measurement data and source-specific models were not available. The study focuses on the Tier 1 and Tier 2 methodologies found in the EEA/EMEP Emission Inventory Guidebook, which are widely accepted and validated methods that ensure methodological transparency and broad applicability. Tier 3 is often more suitable for estimating gaseous emissions such as CO, NO_x, NMVOC, and SO₂. Due to their complexity and lack of comprehensive empirical data, particulate matter emissions are challenging to apply. The computational requirements and resource intensity associated with Tier 3 made it unfeasible, particularly given the goal of ensuring accessibility and reproducibility across different geographic regions with varying data availability.

Considering these constraints, we relied on Tier 1 and Tier 2 methodologies to achieve a balance between accuracy, practicality, and data availability. The emission estimates in this study were reliable and broad in application due to the integration of satellite imagery and established methodologies. Tier 3 may enhance precision, but its feasibility remains dependent on data collection and model refinement advancement.

3. Results

3.1. Research Areas of Satellite Images

The satellite images used in this study were sourced from NASA’s publicly available databases, specifically the MODIS and the VIIRS. The high temporal resolution and ability to identify thermal anomalies associated with active fires make these platforms renowned for their fire detection capabilities [21]. The MODIS satellites provide twice-daily observations with a resolution of 1 km in the thermal band. Against this, VIIRS, onboard Suomi NPP and NOAA-20 satellites, allows for more detailed fire mapping based on 375 m resolution. MODIS and VIIRS imagery provide critical information regarding fire extent, intensity, and temporal variability to evaluate the environmental impacts of agricultural burning [22]. Monitoring fire occurrences, estimating affected areas, and assessing emissions generated by combustion are possible using these platforms that use thermal anomaly detection algorithms. With their ability to cover vast regions over an extended period, they ensure near-real-time fire monitoring, an essential method of analyzing the dynamics of agricultural residue burning and its effects on air quality. The following benefits are provided by MODIS and VIIRS when used in agriculture:

• Frequent Revisit Times: Both systems allow for tracking fire dynamics in near-real time, which is crucial for determining residue combustion patterns over time.
• High Spatial Resolution: Burned areas can be accurately mapped using MODIS’ 1 km resolution and VIIRS’ 375 m resolution.
• Advanced Spectral Capabilities: These features help discriminate between fires and other heat sources, minimizing false detections and ensuring reliable data.

It has been extensively demonstrated in prior research that satellite data can be integrated into agricultural burning studies, demonstrating its utility in quantifying particulates (PM2.5 and PM10) as well as carbon monoxide (CO) and volatile organic compounds (VOCs) [40]. Satellite data are primarily used to identify regions where agricultural practices contribute disproportionately to air pollution, providing valuable information for policymakers to target mitigation strategies [41]. To analyze fire activity associated with agrarian residue burning, this study specifically used MODIS and VIIRS data. Using these datasets, we were able to estimate the environmental and public health impacts of fire events based on their geographic distribution and temporal variability. The satellite images were intentionally selected for analysis to demonstrate the capabilities of the program and system. The selection of this dataset was not intended to provide a comprehensive dataset but rather to highlight the program’s capabilities and its effectiveness. The emphasis on showcasing the system emphasizes its potential for broader environmental monitoring and policy development.

3.2. Emission Estimation by Tier 1 Method

The Tier 1 method was used in this study to estimate air pollution related to particulate matter, including TSP, PM10, and PM2.5. The critical analysis component involved identifying and analyzing areas of interest, particularly those affected by biomass combustion, using satellite imagery. Using NASA’s MODIS and the VIIRS, satellite images were acquired for this study. In order to determine the spatial distribution and extent of emissions, satellite images were rigorously processed. An advanced filter designed to highlight relevant features was applied to the raw satellite data in accordance with a predefined methodology. The filtering process isolates and emphasizes the areas impacted by events such as fires, where biomass combustion contributes significantly to particulate matter emissions. After filtering, dedicated image processing software was used to analyze the filtered images in depth. Software like this is crucial for converting complex image data into a quantitative analysis format. The images are converted into 16-bit files, simplifying the image structure by reducing color and saturation data, which, although visually appealing, may be unnecessary for this specific analytical use. The software detects areas of interest more accurately by analyzing the brightness and intensity distributions of the pixels.

This analysis goes beyond merely identifying affected areas. To separate and thoroughly analyze the images, sophisticated algorithms are used. These algorithms identify distinct areas based on their spectral characteristics. The algorithms can distinguish between various surface types using these spectral signatures and pinpoint areas where biomass combustion occurs. In order to assess dust emissions and their impact on both air quality and the broader environment, it is of primary importance to precisely identify post-fire areas. To mitigate these risks, it is important to identify these regions accurately and quickly. Data processing algorithms and satellite imagery are combined to develop effective mitigation strategies. The results of the batch elements and measurements in California National Park are summarized and presented in

Table 5. Batch analysis results of fire intensity and burned area coverage in California National Park (USA) during the 2018 wildfire season. The recorded satellite images provide insights into fire progression, showing variations in burned land percentages across different dates. The red-marked areas indicate detected fire hotspots, with increasing pixel count corresponding to intensified fire activity over time.

Date	Bitmap of the Scanned Image	Percentage of Fire Coverage, %
15 July 2018		1416 px/0.1335%
22 July 2018		2595 px/0.2447%
29 July 2018		3249 px/0.3063%
12 August 2018		2676 px/0.2523%
19 August 2018		2720 px/0.2523%
30 August 2018		656 px/0.0618

.

The study analyzes particulate emissions from burning crop residues in agricultural fields, utilizing satellite data and Tier 1 methodology. The burned area was determined based on satellite imagery. Agricultural biomass burning and large-scale wildfires contribute significantly to particulate matter emissions. For example, the Carr Fire in Northern California ignited on 23 July 2018 and rapidly expanded, ultimately burning 229,651 acres (92,936 hectares) before it was fully contained on 30 August [42]. The fire emitted significant amounts of PM2.5 and PM10, significantly deteriorating the air quality in California and neighboring regions. We could track the fire progression and estimate its emissions using satellite monitoring. The burned area was assessed and quantified by pollutants using MODIS and VIIRS satellite imagery. There was a peak in fire intensity in late July, coinciding with the largest measured area of biomass combustion. Including wildfire data in this study enhances our understanding of the environmental impact of large-scale biomass combustion. Burning diverse biomass types, such as vegetation and buildings, contributes to higher pollutant concentrations in wildfires than in agricultural burning. In order to design effective fire prevention strategies and emission control measures, it is essential to distinguish between these two types of measures.

The program assessed the extent of fire damage based on data for all cases and specific dates, identifying fires by red dots that marked their locations. According to the scale attached to the image, the measurement field covered 1,060,685 pixels. A resolution of 1535 × 691 corresponds to approximately 101 pixels per 200 km. The bit-depth filter function was used to enhance accuracy and accurately delineate the combustion zone. The use of these processing techniques facilitates image analysis, but they can also introduce limitations. Image scaling and bit-depth conversion simplify visualization but decrease precision, especially when subtle differences in colors or intensities affect measurement results. The interpolated pixel values in point-to-point scaling can also cause distortions, possibly altering fine details while maintaining relative image positioning. In order to avoid errors, measurement artifacts such as gradient loss or distortion of intricate fire patterns must be carefully calibrated. Qualitative assessments must be validated rigorously to be reliable. Burning residues were estimated based on crop stubble remaining post-harvest within the delineated fire-affected area. The product ratio varies according to crop type, though such residues are closely linked to crop biomass. The uncontrolled burning conditions examined in this study contributed to a significant portion of these remnants being consumed, resulting in particulate emissions and air quality impacts.

Satellite data analysis facilitated the determination of the burned area. However, it is crucial to acknowledge that image processing techniques, such as bit-depth conversion and point-to-point scaling, can introduce measurement inaccuracies. Contrary to prior assumptions, crop residues are not equivalent to total biomass but rather represent a percentage thereof. Based on satellite analysis, the burned areas in this study varied from 129,900 hectares to 643,366 hectares across different dates. According to these estimates, total suspended particle (TSP) emissions ranged from 0.754 to 3.732 tons, PM10 emissions ranged from 0.740 to 3.669 tons, and PM2.5 emissions ranged from 0.701 to 3.474 tons. The findings of these studies substantiate that field burning exerts a substantial environmental impact.

The spatial resolutions of MODIS and VIIRS are different, so adjustments were necessary to ensure that the datasets are comparable. Fire signals from MODIS with a 1 km pixel size tend to average over larger areas, underestimating burn areas compared to VIIRS, which detects fires at finer scales. The discrepancy was addressed through cross-validation by aligning MODIS-based estimates with VIIRS-derived data. Because VIIRS detects smaller fires than MODIS, its emission calculation is slightly higher; however, MODIS results show a tendency to smooth fire distribution over larger pixels, which leads to conservative estimates. The mass of burnt residues (Mb) was estimated using a biomass yield coefficient, considering crop-specific residue production factors. Based on the uncontrolled fire scenario used for this study, in which available residues are assumed to be burned, this value approximates biomass production in the study area. In the Tier 1 method, it was assumed that the combustion coefficient (C_f) in this type of burning is close to one, meaning that nearly all available biomass is burned.

Based on MODIS and VIIRS imagery, fires were identified using satellite-derived thermal anomalies, and their burned areas were estimated. Emissions monitored in these cases included total suspended particulate (TSP), PM10, and PM2.5, as these pollutants are key indicators of the impact of biomass burning on air quality. Since Tier 1 uses lower-resolution burned area data, it tends to underestimate emissions from smaller fires. The method lacked accuracy when it came to capturing local fire variations. The study further analyzed factors such as geographic location, crop types, and burning conditions. The calculated values are presented in

Table 6. Calculation of pollutant emissions (E) of particulate matter from field burning events. The values represent total emissions of TSP, PM10, and PM2.5 based on remote sensing data.

Date	Pollution	Burned Area (A), ha	Emission Factor (EFpollutant), kg/kg s.m.	Emission of Pollutants (Epollutant), Tones
15 July 2018	TSP	280,396.04	0.0058 (0.002)	1.626
	PM10		0.0057 (0.002)	1.598
	PM2.5		0.0054 (0.002)	1.514
22 July 2018	TSP	513,861.39	0.0058 (0.002)	2.980
	PM10		0.0057 (0.002)	2.929
	PM2.5		0.0054 (0.002)	2.775
29 July 2018	TSP	643,366.34	0.0058 (0.002)	3.732
	PM10		0.0057 (0.002)	3.669
	PM2.5		0.0054 (0.002)	3.474
12 August 2018	TSP	529,900.99	0.0058 (0.002)	3.073
	PM10		0.0057 (0.002)	3.020
	PM2.5		0.0054 (0.002)	2.861
19 August 2018	TSP	538,613.86	0.0058 (0.002)	3.124
	PM10		0.0057 (0.002)	3.070
	PM2.5		0.0054 (0.002)	2.909
30 August 2018	TSP	129,900.99	0.0058 (0.002)	0.754
	PM10		0.0057 (0.002)	0.740
	PM2.5		0.0054 (0.002)	0.701

.

The area affected by the fires in July and August 2018 underwent dynamic changes, as shown by fire data analysis. The study estimated particulate matter emissions based on the burned area, burn residue mass, burn rate, and pollutant emission factor. Using advanced satellite image analysis technology, satellite data allowed for tracking the fire’s development over time. The burned area was calculated based on MODIS and VIIRS satellite data detecting active fire zones. In order to estimate the total burned land area, satellite imagery was classified and converted into pixels, with a spatial resolution of 70 pixels per 100 km, enabling detailed mapping of combustion zones. An image processing method called bit-depth filtering was used to standardize color representations and improve burned area detection precision. In the estimation of biomass residue mass (Mb), IPCC default residue-to-product ratios were used (e.g., 1.3 for wheat, 1.5 for maize), adjusted using national agricultural statistics. The combustion factors (Cf) were assigned according to EEA and IPCC guidelines: 0.8 for general open field burning, and 0.3–0.5 for incomplete combustion. To ensure consistency with established emission estimation methodologies, these calculations were validated against literature data on biomass burning emissions.

Fire intensity, measured in pixels, increased in July, peaking on 29 July with a value of 3249 pixels before decreasing in August. The study area showed a relatively low proportion of post-fire regions, not exceeding 0.31%. To ensure the accuracy of area measurements, the method for converting pixels to metric units needs validation. Given potential inaccuracies from satellite image processing, such as bit-depth conversion and point-to-point scaling, calibration and validation of methodologies remain essential. Pollutant emissions were calculated in tons by multiplying the following values: burned area (hectares), mass of crop residue burned per hectare, combustion factor, and pollutant emission factor. The study aims to identify the main sources of dust pollution to reduce emissions and improve air quality. Future work may explore smoke dispersion modeling using satellite imagery.

3.3. Emission Estimation by Tier 2 Method

In order to obtain a more accurate estimate of the emissions from crop residue combustion, this study applied a Tier 2 methodology. Comparing Tier 2 methodology to Tier 1, it uses a more detailed approach considering inputs specific to a particular region or crop. Satellite imagery was used for Tier 2 analysis to identify key parameters for calculating emissions. It was considered that such areas as Arkansas, Heilongjiang Province in China, and India have diverse agricultural and climatic characteristics. These locations allow us to assess the applicability of the Tier 2 method in different geographic contexts and residual management systems. These sections describe the adaptation and implementation of Tier 2 methodology using satellite data for each region.

The Tier 2 method was used in regions where it was possible to gather detailed biomass combustion parameters, allowing for a more accurate assessment of emissions. To enhance the accuracy of the burn area-to-emission conversion, region-specific emission factors were used to estimate TSP, PM10, and PM2.5. The results of Tier 2 confirmed its ability to estimate localized emissions more accurately than Tier 1 in areas with high fire intensity. In comparison to the average yields of cereal crops, a key crop globally, cereal yields vary significantly by species, forming ranges of 4.5–5.5 tonnes/ha for wheat, 3.5–4.5 tonnes/ha for barley, 6.5–7.5 tonnes/ha for corn, and 5.0–6.0 tonnes/ha for rice [43]. The values representing agricultural productivity also determine the amount of crop residues generated, which may be burned in agricultural practice [44]. To analyze this process’s impact on air quality, the present study developed four pollution models for the post-crop fields of the aforementioned cereal species, like wheat, barley, corn, and rice.

The critical parameter in assessing emissions from agricultural residue burning is the combustion factor (C_f), which represents the proportion of biomass actually consumed during a burning event [37]. The Intergovernmental Panel on Climate Change (IPCC) provides guidance on Cf values, which are not fixed and depend on factors such as biomass type and regional burning practices. Broad assessments of agricultural burning often employ a default C_f of 0.8 (80% combustion), as suggested by some IPCC reports [45]. However, for more refined estimations, particularly when dealing with residues from crops like wheat and maize in certain regions, a lower C_f of 0.3 may be more appropriate.

The relative contribution of crop residues to total biomass burning also varies geographically. In China [46], for instance, crop residues are estimated to account for approximately 60% of the total biomass burned nationally. However, this proportion can fluctuate significantly across different regions [8]. Specific crop types exhibit varying combustion characteristics; rice straw, for example, can have a combustion completeness of up to 80% [47]. Based on literature data defining the biomass content of crops, the combustion factor (Cf) for burning crop residues was assumed to depend on a measured area. Further research will focus on analyzing the amount of post-crop biomass for crops studied by the IPCC [45].

3.3.1. Screen Estimation for Wheat Residue Burnouts in India

Stubble burning remains a common and problematic agricultural practice in northern India, particularly in the Punjab, Haryana, and Uttar Pradesh states [48]. Farmers frequently burn residue in preparation for wheat planting as a quick and low-cost method [8]. There are two main periods when seasonal burning intensifies: October–November, just after the rice harvest, and April–May, just after the wheat harvest. This study analyzed the burning process of wheat residues between April and May in order to design a model [49]. Satellite data will be used to analyze stubble-burning areas and assess their impact on emissions as a result of this study. An inventory of agricultural emissions, including those from biomass burning, will be conducted using the IPCC methodology. Satellite data and IPCC methodology will provide an objective and spatially differentiated assessment of burned fields in key seasons. Analyzing the collected data will enable a regional assessment of particulate matter (TSP, PM2.5, PM10), carbon monoxide, and aromatic hydrocarbon emissions caused by biomass burning. The results of the batch elements and measurements of wheat fire in India are summarized and presented in

Table 7. Spatial distribution of wheat stubble burning in India (April–May 2021). This dataset presents fire (red color areas indicate active fire zones) coverage percentages detected via satellite images and quantifies the seasonal fluctuations in field burning events.

Date	Bitmap of the Scanned Image	Percentage of Fire Coverage, %
5 April 2021		16,146 px/ 1.5222%
12 April 2021		10,934 px/ 1.0308%
19 April 2021		7674 px/ 0.7235%
26 April 2021		14,468 px/ 1.364%
3 May 2021		13,570 px/ 1.2794%
10 May 2021		9565 px/ 0.9018%
17 May 2021		3313 px/ 0.3123%
05-24-2021		1850 px/ 0.1744%

.

Data from all examined instances and specific time points were processed to identify fire-affected zones. Fire occurrences were visually represented as a spatial distribution using red indicators overlaid on the imagery. Based on the image’s integrated scale, the total measurement domain was quantified as 1,060,685 pixels (with a resolution of 1535 × 691 px). This pixel count corresponded to a spatial scale of 70 pixels per 100 km, indicating the potential for detailed measurements of combustion zones. Bit-depth filtering was employed to standardize color representation and precisely determine the actual burned area. Acknowledging that both the bit-depth conversion process and image rescaling via the point-to-point method inherently introduce potential measurement errors is crucial. This potential for accuracy loss becomes particularly significant in scenarios where subtle color variations or intensity gradients are critical for precise interpretation.

As a result of the precise measurement of burn area derived from pixel counts and spatial scaling, it was possible to estimate and analyze particulate matter emissions associated with these fire incidents. This study summarizes its findings by precisely quantifying particulate emissions. The IPCC’s Tier 2 methodology was applied to achieve high estimation accuracy, incorporating detailed regional data on wheat characteristic emission factors and, most importantly, combustion area extent. The remainder of the study will present a detailed characterization of particulate emissions (TSP, PM2.5, and PM10) generated by burning the post-crop biomass of these key crops, considering spatial and seasonal variations. The results of emissions of pollutants for wheat were calculated and are presented in

Table 8. Estimation of wheat stubble burning emissions in India (April–May 2021). Emissions of TSP, PM10, and PM2.5 are calculated based on detected burned areas, emission factors, and biomass residue parameters. The highest emissions were recorded during peak burning periods.

Date	Pollution	Burned Area (A), ha	Emission Factor (EFpollutant), kg/kg s.m.	Emission of Pollutants (Epollutant), Tones
5 May 2021	TSP	2,306,571.43	0.0058 (0.002)	13.378
	PM10		0.0057 (0.002)	13.147
	PM2.5		0.0054 (0.002)	12.455
12 April 2021	TSP	1,562,001.23	0.0058 (0.002)	9.060
	PM10		0.0057 (0.002)	8.903
	PM2.5		0.0054 (0.002)	8.435
19 April 2021	TSP	1,096,285.71	0.0058 (0.002)	6.358
	PM10		0.0057 (0.002)	6.249
	PM2.5		0.0054 (0.002)	5.920
26 April 2021	TSP	2,066,857.14	0.0058 (0.002)	11.988
	PM10		0.0057 (0.002)	11.781
	PM2.5		0.0054 (0.002)	11.161
3 May 2021	TSP	1,938,571.43	0.0058 (0.002)	11.244
	PM10		0.0057 (0.002)	11.050
	PM2.5		0.0054 (0.002)	10.468
10 May 2021	TSP	1,366,428.57	0.0058 (0.002)	7.925
	PM10		0.0057 (0.002)	7.789
	PM2.5		0.0054 (0.002)	7.379
17 May 2021	TSP	473,285.72	0.0058 (0.002)	2.745
	PM10		0.0057 (0.002)	2.698
	PM2.5		0.0054 (0.002)	2.556
24 May 2021	TSP	264,285.74	0.0058 (0.002)	1.533
	PM10		0.0057 (0.002)	1.506
	PM2.5		0.0054 (0.002)	1.427

.

The study presents a detailed assessment of particulate matter emissions from post-harvest biomass burning in India. Total suspended particle (TSP), PM10, and PM2.5 emissions exhibit significant seasonal variations, peaking during intensive field burning activities in early and late April. Approximately 13.378 tons of TSP, 13.147 tons of PM10, and 12.455 tons of PM2.5 were emitted on 5 April 2021. It was reported that on 26 April 2021, PM10 emissions fell to 11.781 tons, while PM2.5 emissions fell to 11.161 tons. The lowest emissions occurred on 24 May 2021, coincident with the smallest area burned. The reduction in agricultural burning resulted in TSP emissions of 1.533 tons, PM10 emissions of 1.506 tons, and PM2.5 emissions of 1.427 tons. Particularly between April and May, crop residue burning contributes significantly to particulate matter pollution. Therefore, a stricter regulatory regime and improved residue management strategies are urgently needed to limit agricultural emissions. There is a significant risk to human health and air quality from PM2.5 and PM10.

3.3.2. Screen Estimation for Barley Residue Burnouts in India

This study aims to investigate the widespread practice of burning agricultural residues and their significant environmental impact. A specific focus will be placed on barley residues. Field burning releases significant amounts of pollutants, including particulate matter, which pose health risks and have a negative impact on regional air quality. To analyze the extent of burned areas and quantify their effect on emissions, this research uses satellite data and IPCC methodology to develop a model for spatially analyzing this practice. The study aims to identify specific areas with high emissions to maximize sustainability and minimize the adverse environmental effects of residue burning. The results of the batch elements and measurements of wheat fire in India are summarized and presented in

Table 9. Spatial distribution of barley residue burning in India (October–November 2024). The fire (red color areas indicate active fire zones) coverage percentages were derived from satellite image processing, illustrating the progressive increase in burned area throughout the season.

Date	Bitmap of the Scanned Image	Percentage of Fire Coverage, %
30 September 2024		3295 px/ 0.3106%
7 October 2024		7197 px/ 0.6785%
14 October 2024		9258 px/ 0.8728%
21 October 2024		10,019 px/ 0.9446%
27 October 2024		18,208 px/ 1.7166%
3 November 2024		17,685 px/ 1.6673%
10 November 2024		14,877 px/ 1.4026%
17 November 2024		24,575 px/ 2.3169%

.

This study analyzes the significant environmental impacts of widespread barley residue burning. Based on the image’s integrated scale, the total measurement domain was quantified as 1,060,685 pixels (with a resolution of 1535 × 691 px). This analysis used 68 pixels, the equivalent of 100 km in reality, for the map scale. The research employs advanced techniques to accurately quantify these emissions and provide a spatially detailed assessment of the practice, ultimately emphasizing the urgent need to transition to more sustainable agricultural practices. The results of emissions of pollutants for wheat were calculated and are presented in

Table 10. Emissions of particulate pollutants from barley residue burning (October–November 2024). This dataset provides a quantitative assessment of TSP, PM10, and PM2.5 emissions and highlights the peak burning periods that contribute to significant air quality deterioration.

Date	Pollution	Burned Area (A), ha	Emission Factor (EFpollutant), kg/kg s.m.	Emission of Pollutants (Epollutant), Tones
30 September 2024	TSP	484,558.82	0.0078 (0.001)	3.780
	PM10		0.0077 (0.001)	3.731
	PM2.5		0.0074 (0.001)	3.586
7 October 2024	TSP	1,058,382.353	0.0078 (0.001)	8.255
	PM10		0.0077 (0.001)	8.150
	PM2.5		0.0074 (0.001)	7.832
14 October 2024	TSP	1,361,470.588	0.0078 (0.001)	10.619
	PM10		0.0077 (0.001)	10.483
	PM2.5		0.0074 (0.001)	10.075
21 October 2024	TSP	1,473,382.353	0.0078 (0.001)	11.492
	PM10		0.0077 (0.001)	11.345
	PM2.5		0.0074 (0.001)	10.903
27 October 2024	TSP	2,677,647.059	0.0078 (0.001)	20.886
	PM10		0.0077 (0.001)	20.618
	PM2.5		0.0074 (0.001)	19.815
3 November 2024	TSP	2,600,735.294	0.0078 (0.001)	20.286
	PM10		0.0077 (0.001)	20.026
	PM2.5		0.0074 (0.001)	19.245
10 November 2024	TSP	2,187,794.118	0.0078 (0.001)	17.065
	PM10		0.0077 (0.001)	16.846
	PM2.5		0.0074 (0.001)	16.190
17 November 2024	TSP	3,613,970.588	0.0078 (0.001)	28.189
	PM10		0.0077 (0.001)	27.828
	PM2.5		0.0074 (0.001)	26.743

.

The table presents data on emissions of total suspended particulates (TSPs), PM10, and PM2.5 from agricultural burning between 30 September and 17 November 2024. A direct correlation exists between emissions and the size of the burned area. The highest emissions were observed on 17 November 2024, with 3,613,970 ha burned, resulting in 28.189 tons of TSP, 27.828 tons of PM10, and 26.743 tons of PM2.5 emitted. Conversely, the lowest emissions occurred on 30 September 2024, with a significantly smaller burned area of 484,559 ha, resulting in 3.780 tons of TSP, 3.731 tons of PM10, and 3.586 tons of PM2.5 emitted. A gradual increase in emissions is evident from early October, peaking in late October and November, indicating intensified agricultural burning activities during this period. These findings highlight the seasonal nature of biomass burning, a significant contributor to particulate pollution, with PM2.5 posing the most serious risk to air quality and human health. The results underscore the urgent need for effective mitigation strategies, including stricter regulations and alternative crop residue management techniques, to reduce air pollution from open-field burning.

3.3.3. Screen Estimation for Maze Residue Burnouts in Arkansas (USA)

There are also agricultural regions in the United States where burning crop residue is a common practice. Arkansas farmland provides another clear illustration of this phenomenon. Stubble burning is a common agricultural practice in Arkansas, especially in the Delta region, characterized by rice, corn, soybean, and wheat cultivation. Following the harvest season, farmers ignite crop residues to efficiently clear fields and prepare the land for subsequent planting. Satellite imagery analysis from October 2024 reveals substantial fire activity across northeastern Arkansas, indicated by thermal anomalies and visible smoke plumes. While this method offers a cost-effective alternative to mechanical residue management, it significantly contributes to air pollution and greenhouse gas emissions. The release of particulate matter (including TSP, PM2.5, and PM10) from these fires poses a local respiratory health risk. Recent legislative initiatives have aimed to more strictly regulate field burning, proposing enhancements to air quality monitoring and adopting alternative residue management strategies. Remote sensing data are valuable for quantifying emissions, assessing environmental impacts, and informing policy development to foster sustainable agricultural practices. The results of the batch elements and measurements of maze fire in Arkansas (USA) are summarized and presented in

Table 11. Fire (red color areas indicate active fire zones) coverage estimation for maize stubble burning in Arkansas (USA) during October 2024. This analysis utilizes remote sensing data to map burned areas and identify seasonal burning patterns.

Date	Bitmap of the Scanned Image	Percentage of Fire Coverage, %
18 October 2024		2645 px/0.2494%
21 October 2024		3042 px/0.2868%
24 October 2024		2954 px/0.2785%
29 October 2024		2804 px/0.2644%

.

An analysis of satellite data from selected time periods in Arkansas was conducted to identify areas affected by cornfield burning. In order to provide reliable emissions estimates, the analysis had to delineate the area of post-crop biomass burning precisely. As a preliminary step in the research methodology, satellite imagery was used to interpret and identify potential burn areas. Thermal anomalies and smoke plumes were detected based on characteristic spectral signatures, signaling active fires and biomass-burning areas. The following step was to measure precisely the areas that had been tentatively identified as burned. An integrated scale of satellite imagery was used, including information about the image’s spatial representation. Based on this scale, the image has a total measurement area of 1,060,685 pixels, with a resolution of 1535 × 691 px. The spatial resolution was defined by 119 pixels in the image, corresponding to an area of 20 km in the field.

Using spatial resolution, counting pixels classified as burned areas to measure burned areas of corn fields precisely and in detail was possible. The emission characteristics of corn and, most importantly, the extent of the combustion area are based on regional data on corn emission factors. Spatial and seasonal variations will be considered after a detailed characterization of particulate emissions (TSP, PM2.5, and PM10) generated from burning the post-crop biomass of these key crops. The results of emissions of pollutants for maize were calculated and are presented in

Table 12. Estimated emissions from maize stubble burning in Arkansas (USA). Emission estimates for TSP, PM10, and PM2.5 are based on burned area calculations and biomass combustion factors, providing critical insights into the impact of agricultural burning on regional air quality.

Date	Pollution	Burned Area (A), ha	Emission Factor (EFpollutant), kg/kg s.m.	Emission of Pollutants (Epollutant), Tones
18 October 2024	TSP	44,453.78	0.006 (0.002)	0.267
	PM10		0.0062 (0.002)	0.276
	PM2.5		0.006 (0.002)	0.267
21 October 2024	TSP	51,126.05	0.006 (0.002)	0.307
	PM10		0.0062 (0.002)	0.317
	PM2.5		0.006 (0.002)	0.298
24 October 2024	TSP	49,647.06	0.006 (0.002)	0.308
	PM10		0.0062 (0.002)	0.298
	PM2.5		0.006 (0.002)	0.283
29 October 2024	TSP	47,126.05	0.006 (0.002)	0.292
	PM10		0.0062 (0.002)	0.283
	PM2.5		0.006 (0.002)	0.307

.

This study presents the findings of an analysis examining pollutant emissions from the combustion of maize residues. Measurements of particulate matter emissions were conducted on four distinct dates in October 2024. The spatial extent of the fires ranged from 44,453.78 to 51,126.05 hectares. Emission factors for TSP and PM2.5 were determined to be 0.006 kg/kg, while the emission factor for PM10 was slightly higher at 0.0062 kg/kg. Total emissions ranged from 0.267 to 0.317 tons, notably lower than previous estimations. The highest emission level, reaching 0.317 tons of PM10, was recorded on 21 October 2024. These results underscore the necessity for further investigation into the environmental consequences of maize residue burning and the effectiveness of residue management strategies in mitigating atmospheric pollution.

3.3.4. Screen Estimation for Rice Residue Burnouts in China

Heilongjiang in northeastern China is one of the most prominent agricultural provinces that burns crop residues after harvest to prepare fields for planting in the spring. Satellite data analyses have documented numerous thermal anomalies and smoke plumes, especially during springtime. The prevalence of spring burning since 2015 is demonstrably linked to implementing restrictions on autumn burning [46,50]. A Tier 2 methodology is applied to calculate pollutant emissions from this agricultural practice, which accounts for the extensive combustion of post-crop biomass. This burning practice is not exclusive to China. Satellite observations also record numerous wildfires along the Amur River on the Russian side of the border [32]. These fires are likely associated with similar agricultural residue management techniques and the subsequent fire spread into forests and grasslands. Residues from staple crops are the primary fuel source for this widespread combustion. Satellite observations indicate that the most intense fire activity is frequently concentrated in Heilongjiang’s central plains, a region characterized by vast, contiguous agricultural fields. This extensive and spatially concentrated burning significantly contributes to regional air pollution, notably impacting downwind urban centers. The results of the batch elements and measurements of rice fire in Heilongjiang are summarized and presented in

Table 13. Analysis of rice field fire extent in Heilongjiang, China, using remote sensing techniques. The table presents the scanned image pixel count and the corresponding percentage of fire-affected areas (red color areas indicate active fire zones) detected between 19 April and 21 April 2021. Satellite observations identified concentrated fire activity in Heilongjiang’s central plains, significantly contributing to regional air pollution.

Date	Bitmap of the Scanned Image	Percentage of Fire Coverage, %
19 April 2021		12,029 px/1.1341%
20 April 2021		19,899 px/1.8761%
21 April 2021		26,923 px/2.5383%

.

To accurately estimate the extent of post-harvest biomass burning in key agricultural regions of China, satellite data analysis methodologies were employed to identify and precisely measure affected areas. The spatial resolution analysis established a relationship wherein 96 pixels on the image corresponded to a ground distance of 100 km. Analysis of the integrated scale of the satellite imagery determined a total measurement area of 1,060,685 pixels and a resolution of 1535 × 691 px. This pixel size enabled detailed measurements of burn areas with sufficient spatial accuracy for subsequent analyses. The research methodology commenced with the visual interpretation of satellite imagery as an initial step in identifying potential wildfire outbreaks.

During this stage, characteristic spectral signatures were analyzed, enabling the detection of thermal anomalies and smoke plumes as key indicators of active fires and plant biomass combustion. Subsequently, the surface area of regions tentatively identified as burned was precisely measured. Based on the satellite image scale and spatial resolution, pixel counting was utilized to quantify the burned areas. The data acquired regarding burn areas then served as the foundation for the subsequent analysis stage, which involved utilizing the measurement results and the Tier 2 methodology to calculate particulate matter emissions. Data on rice residues burning out were collected in regional areas, allowing detailed characterization of particulate emissions originating from burning post-crop biomass. The results of emissions of pollutants for maize were calculated and are presented in

Table 14. The results of emissions of pollutants for maize.

Date	Pollution	Burned Area (A), ha	Emission Factor (EFpollutant), kg/kg s.m.	Emission of Pollutants (Epollutant), Tones
19 April 2021	TSP	1,253,020.83	0.0058 (0.003)	7.268
	PM10		0.0058 (0.003)	7.268
	PM2.5		0.0055 (0.003)	6.892
20 April 2021	TSP	2,072,812.50	0.0058 (0.003)	12.022
	PM10		0.0058 (0.003)	12.022
	PM2.5		0.0055 (0.003)	11.400
21 April 2021	TSP	2,804,479.17	0.0058 (0.003)	16.266
	PM10		0.0058 (0.003)	16.266
	PM2.5		0.0055 (0.003)	15.425

.

The table presents the results of calculations concerning total suspended particles (TSP), PM10, and PM2.5 from burning maize residue between 19 April and 21 April 2021. This period saw an increase in burned area from 1,253,020.83 ha to 2,804,479.17 ha, reflecting the growing scale of biomass combustion. In this study, the emission factors for each pollutant were 0.0058 kg/kg for TSP and PM10 and 0.0055 kg/kg for PM2.5. Therefore, particulate emissions increased with burn area. On the first day of the experiment, TSP and PM10 emissions amounted to 7.268 tons and PM2.5 emissions were 6.892 tons, but by the third day, TSP and PM10 emissions had risen to 16.266 tons and PM2.5 emissions to 15.425 tons. Maize residue burning significantly impacts air quality, highlighting the need to consider alternative methods of post-harvest residue management. Additionally, a margin of error of 0.003 kg/kg dry matter should be considered, which may affect the accuracy of emission calculations. Further research is needed to better understand emissions’ spatial and temporal variability and the effectiveness of different air pollution reduction strategies.

4. Discussion

The results of the conducted study confirm that the combustion of agricultural biomass significantly affects the emission of suspended particulate matter, especially PM2.5 and PM10. The analysis of satellite images allowed for the precise identification of areas affected by fires and the quantification of emitted pollutants, making this method a valuable tool for monitoring and reducing particulate emissions. The obtained results are consistent with previous studies, including Roman et al. (2021) [1], who demonstrated the spatial variability of particulate emissions related to agricultural activities in Poland. Similar observations were made by Gokul et al. (2023) [2] and Awasthi et al. (2010) [43], emphasizing the harmful effects of particulate emissions from biomass combustion on human health. Harrison et al., (2004) [3] pointed out differences in the chemical composition of PM10 and PM2.5 under various environmental conditions, which could serve as a basis for further research into the specific properties of particles generated by the combustion of crop residues. The high-resolution satellite data and precision image processing techniques used in this study allow for more accurate estimation of emissions under different environmental conditions than previous studies.

Compared to studies conducted in other parts of the world, the applied methods have proven effective in various geographical contexts. For example, studies by Jain et al., (2014) [8] in India and Li et al., (2016) [32] in China confirm the seasonal nature of particulate emissions from burning crop residues. Meanwhile, Singh et al. (2020) [26] demonstrated that using thermal anomalies from VIIRS enables precise emission mapping from agricultural fields in the United States. Differences in emissions between regions primarily result from variations in crop structure, combustion rates, and local agricultural practices, highlighting the need for further calibration of emission models.

The effects of spatial resolution differences on fire detection and emission estimation highlight the challenges associated with multi-resolution datasets. Using VIIRS, burned areas can be pinpointed with greater precision, allowing small-scale fires to be captured that MODIS might overlook. Consequently, VIIRS may detect transient, low-intensity fire events with little effect on emissions, so additional filtering may be needed. The MODIS satellite, on the other hand, provides a long-term, consistent global dataset but looks at emissions from fragmented fire activity, which could be underestimated. Using a correction factor applied to MODIS data, the results were aligned better with those from VIIRS. Refining resolution-based adjustments to improve emission estimation across different satellite datasets will be the focus of future research.

The results indicate that a method’s choice significantly influences emission estimates. Because Tier 1 uses generalized emission factors, it underestimates emissions from smaller fires. Tier 2 produced more accurate, and therefore higher, emission values, proving it is suitable for detailed regional assessments with more localized input parameters. In regions with limited access to regional data, its applicability is limited. A trade-off exists between a model’s broad applicability (Tier 1) and its precision (Tier 2). Despite the numerous advantages of the applied methods, certain study limitations should be considered. The analysis relies on satellite data, which may be subject to errors due to atmospheric conditions such as cloud cover. Additionally, the adopted emission factors are averaged and may not reflect actual values for specific combustion conditions. Moreover, the study does not account for the direct health effects associated with exposure to suspended particulate matter, representing a crucial area for further research.

The results clearly indicate that burning crop residues remains a major source of suspended particulate emissions in agriculture. Its reduction requires the implementation of effective regulations, such as Directive 2008/96/EC [51] of the European Parliament, as well as using modern monitoring technologies and alternative methods for managing crop residues. In future research, it will be essential to improve emission estimation methods, evaluate the effectiveness of alternative agricultural practices, and develop real-time monitoring tools, enabling a better understanding of the scale and impact of emissions from agricultural biomass.

5. Conclusions

The study’s findings offer valuable insights that can shape future research. The originality of the results lies in the comprehensive assessment of the environmental and health impacts associated with field burning, particularly concerning particulate emissions and their dispersion over long distances. The study uses high-resolution satellite data and advanced analytical methods to assess fire-related particulate emissions accurately. Among the important research gaps highlighted in this study are the need for a more dynamic analysis of emission patterns, the influence of meteorological conditions on pollutant transport, and the effectiveness of alternative residue management approaches.

The practice of field burning in agriculture has profound environmental and health consequences. This method significantly contributes to wind erosion and releases large quantities of PM10 and PM2.5 particulate matter, which are major air pollutants. While burning plant residues temporarily enhances soil fertility by releasing nutrients, it also causes severe adverse effects such as biodiversity loss, soil structure degradation, and the emission of greenhouse gases and harmful dust particles. PM10 particles are generally removed from the air through sedimentation within a few hours. However, finer PM2.5 particles can remain suspended for days or weeks, traveling up to 2500 km from their origin. These fine particles pose serious health risks, particularly to the respiratory and cardiovascular systems.

The studies should aim to refine emission estimation models using real-time satellite observations, improve meteorological corrections, and examine the long-term effects of field burning on regional and global air quality. Investigation is needed to calculate any differences in toxicity of particulate matter emitted under different combustion conditions. Adopting alternative crop residue management practices and enforcing stricter regulations are essential to mitigate the negative consequences of field burning. Integrating advanced monitoring technologies with strong environmental policies can effectively reduce its harmful impacts and improve overall air quality.

The use of advanced waste management techniques, including biochar production, composting, and precision-controlled soil incorporation, could reduce emissions from agricultural burning. Future research should investigate these technologies and evaluate their feasibility in various agricultural contexts. The future of agricultural burning emission management should include evaluating advanced waste management techniques, such as biochar production, composting, and precision-controlled soil incorporation. Moreover, interdisciplinary approaches that combine atmospheric science, remote sensing, and policy studies are necessary to create effective mitigation strategies. By addressing these gaps, future studies can provide more precise recommendations for reducing agricultural air pollution and its associated health risks.