Environmental studies using remote sensing methods

Abstract:

In this article, we will tell you about one of the most advanced methods in ecological research – remote sensing of the Earth.

Remote sensing methods enable us to obtain information about the state and dynamics of ecosystems, identify changes in vegetation, water resources, soil, and other elements of the environment. They also contribute to assessing biodiversity and detecting environmental threats. All of this helps us better understand ecological processes, identify trends, and make informed decisions in the field of environmental protection and sustainable development.

Introduction

In the modern world, environmental issues are one of the main challenges facing humanity. Timely and effective research methods are necessary to understand and assess the state of the environment and its changes. In recent decades, remote sensing of the Earth has become an important tool for ecological research, allowing valuable data on ecosystem conditions and changes to be obtained with high spatial and temporal resolution.

Main objectives of ecological research:

• Monitoring the state of the environment;
• Generating information about the current state of the external environment and the degree of its pollution;
• Conducting large-scale observations of anthropogenic impact factors;
• Forecasting trends in the changing state of the natural environment;
• Identifying sources of anthropogenic influence and organizing observation of them;
• Forecasting and controlling natural disasters and technological accidents.

Reasons for Research

Today, our planet faces numerous threats to the environment, some of which are local, while others are common to all countries. Environmental research helps to anticipate and prevent them.

Evaluation of Environmental Conditions: Environmental research allows for the study and analysis of the condition of natural ecosystems, including land, water, air, and biodiversity. They help to determine the level of pollution, climate change, degradation of natural resources, and other aspects that may impact ecosystems.

Monitoring and Forecasting Changes: Environmental research enables the tracking of long-term changes in the environment and its components. This helps in taking measures for conservation and sustainable use of natural resources.

Assessment of Environmental Risk: Environmental research helps assess the risks associated with specific activities or pollutants.

Development and Evaluation of Environmental Policies and Programs: Environmental research provides scientific basis for decision-making aimed at conserving natural resources, protecting biodiversity, and reducing negative impacts on the environment.

Management and Restoration of Ecosystems: Environmental research contributes to understanding the processes occurring in natural ecosystems and helps develop effective methods for their restoration and conservation.

Thus, environmental research plays a crucial role in the assessment, understanding, and preservation of the environment, as well as in the development of sustainable approaches to its use and management.

Fig.1. Forest loss monitoring. Rondonia, Brazil, comparing images from 2000 and 2013, with the difference indicating forest loss shown in red on the 3rd panel.

Directions for Obtaining Environmental Data

To study the state of ecology in any part of the Earth, the following data are required:

• Pollution or deviations in vegetation, land, forests, and water resources, posing a threat to human and animal habitats.
• Atmospheric harmful chemical molecular substances, causing health disruptions to humans and negatively impacting the overall condition of the Earth's biosphere.
• Differences in living in cities, large agglomerations, or rural areas; the above data require separate approaches to ecosystem assessment and obtaining localized data based on the place of residence.
• Statistical, mathematical, and indicative indicators for decision-making in the restoration of the biosphere and biodiversity.
• Animal life on land, in the air, and in water. Microorganisms and their sphere of influence.
• Industrial negative human impact.

Consequences

The absence of environmental monitoring can have serious consequences for the environment and human well-being:

Deterioration of Environmental Quality: Without monitoring, uncontrolled pollution and degradation of the environment can lead to deterioration of air, water, and soil quality. This can negatively impact the health of humans, animals, and plants, as well as lead to loss of biodiversity.

Unbalanced Use of Natural Resources: Without monitoring, it is impossible to assess and control the level of their use. This can lead to the loss of ecosystems.

Threat to Biodiversity: Without monitoring, it is impossible to track changes in biodiversity and preserve vulnerable species and ecosystems. Loss of biodiversity will disrupt ecological balance, worsen ecosystem conditions, and lead to the loss of valuable natural resources.

Risk of Environmental Disasters: Without systematic monitoring, it is difficult to detect and predict possible environmental disasters.

Lack of Scientific Basis for Decision Making: Environmental monitoring provides the scientific data needed for making informed decisions in environmental protection and policy development.

Fig.2. Forecasts of mussel growth (left panel) in the Bay of Mont Saint-Michel, France, using chlorophyll-a μg/l, using Sea-WIFS sea surface temperature data as input to a bioenergetics model predicting growth patterns (right panel)

Therefore, environmental monitoring is an important tool for assessing and controlling the state of the environment, maintaining its stability, and taking effective measures to protect nature and preserve biodiversity.

Why Remote Sensors Are at the Forefront of Environmental Research Today?

A powerful and scalable way to obtain data for understanding, managing, and monitoring ecosystem restoration is through remote sensing methods.

Remote sensing obtains information by using electromagnetic radiation reflected or emitted by the Earth's surface and atmosphere, or the devices emit electromagnetic radiation themselves and register the reflected signal – this is an active sensing method. Direct use of remote sensing to assess vegetation phenology (study of periodic life cycle events of plants in relation to the environment) has been conducted for many years.

Advantages of remote methods compared to traditional ones:

• Voluminous information and wide coverage of territories. Satellites capture areas of tens to hundreds of kilometers, with high spatial and spectral resolution.
• Possibility of monitoring on various time scales. Large volumes of archived data allow for research covering weeks or years, as well as observing short-term evolutionary processes.
• Non-contact nature of research. Remote sensing does not require direct contact with the objects or ecosystems under study.
• Multi-spectral data for analyzing various aspects of ecosystems. Remote sensing allows collecting data in different spectral ranges, including visible, infrared, and microwave spectra.
• Possibility of detecting changes and trends.

Disadvantages of remote methods:

• Lack of direct ground information or confirmation.
• Limitations in data resolution and accuracy. Although many satellites have high spatial resolution of 1 meter, in some cases, especially when studying small-scale objects, there may be limitations and the need for additional analysis.
• Need for careful calibration and data validation.

It is also important to understand that environmental monitoring and research are simultaneously close and different. Regarding the difference between environmental monitoring and environmental research:

• Environmental monitoring covers long-term and systematic observation of environmental parameters, establishment of long-term trends, and monitoring the state of natural resources. Monitoring is often conducted after incidents of various environmental damage to assess the scale and volume of corrective actions.
• Environmental research encompasses a narrower area of study aimed at obtaining new knowledge about specific aspects of ecosystems, processes, or impacts. The main goals of research are to prevent impending disasters.

Fig.3. Monitoring of oil spill in the Gulf of Mexico a month after the explosion on the Deepwater Horizon oil platform. Fragment of MODIS Terra image (composite of channels 1, 4, 3). The image was obtained on 17.05.2010 at 16:40 UTC, with a resolution of 250 m. Sun glint zone. It can be clearly seen how the "jet" under the action, most likely, of the current, rushed southeast towards Cuba.

Methods

Some of the main methods of Earth remote sensing used in ecological research include:

Multispectral Imaging:

Multispectral imaging captures image data in specific wavelength ranges of the electromagnetic spectrum. Wavelengths can be separated by filters or detected using instruments sensitive to certain wavelengths, including light frequencies beyond the visible range, i.e., infrared and ultraviolet. This can allow extracting additional information that the human eye cannot perceive with its visible receptors for red, green, and blue colors.

Multispectral satellites collect data in 5–20 spectral ranges. The image usually consists of three primary colors and several infrared parts of the spectrum. This makes them more cost-effective.

The advantages of this method include a wide spectral coverage, allowing for information gathering on various Earth surface characteristics. Comparatively low data cost and availability due to the presence of mass-produced satellite systems. Large data volume and the ability to conduct large-scale studies.

Fig.4. Example of multispectral imaging of fields using the DJI Phantom 4 Multispectral UAV.

Hyperspectral Imaging

Hyperspectral remote sensing differs from classical optical multispectral approaches in the number and width of spectral bands recorded by the sensor. While multispectral sensors measure the optical part of the electromagnetic spectrum in several (< 10) relatively broad bands (approximately 40–100 nm wide), hyperspectral sensors measure the optical part of the electromagnetic spectrum in many (> 50, often many more) regularly spaced narrow bands approximately 5 to 10 nm wide. Hence, hyperspectral data allow for detailed selection of all spectral reflection characteristics that can be used for quantitative assessment of various biological, geological, and chemical properties of the Earth's surface.

It is interesting to note that classical multispectral sensor systems are also increasingly using narrow spectral bands to measure specific vegetation properties. Multispectral instruments on board Sentinel-2a and Sentinel-2b, for example, have three narrow bands in the near-red spectral region, as well as a narrow band in the near-infrared region, allowing for a more detailed characterization of vegetation properties.

Currently, several spaceborne hyperspectral sensors are available (e.g., Hyperion on board EO-1 or the Compact High Resolution Imaging Spectrometer on board PROBA-1). These spaceborne hyperspectral sensors will help better understand the global spatial and temporal variability of ecosystem biogeochemical properties.

Lidar Scanning

Optical remote sensing is limited in studying vegetation cover due to the inability of sunlight to penetrate through tree canopies and other vegetation, leading to the emergence of a new type of active sensor - LiDAR. It allows for more precise and detailed study of vegetation structure. LiDAR sensors use laser beams to measure the time required for the laser beam to refract and reflect back to the sensor. Based on this data, the height above the ground can be determined. By creating a three-dimensional point cloud from reflected laser beams, maximum height values can be identified and converted into a digital surface model - a regular grid of cells reflecting vegetation height.

However, despite the capabilities of airborne LiDAR for understanding spatial (and spatio-temporal) dynamics of ecosystems, it remains an expensive tool due to the high cost of operating platforms such as UAVs, airplanes, and helicopters, especially over large spatial extents. Thus, analysis often is limited to areas with existing data or small areas and individual measurements. This gap can be filled by new spaceborne LiDAR data from the Global Ecosystem Dynamics Investigation (GEDI) system, which provides "LiDAR footprints" with a diameter of 25 m on a regular 60 × 600 m grid. By merging GEDI with optical satellite data in the future, global maps of key vegetation features, such as global tree height, can be created.

Fig.6. Examples of LiDAR point clouds for a research forest plot of 500 m ² (left panel) with individual trees segmented to obtain structural metrics based on trees (e.g., height; middle panel). Data were collected from a helicopter in 2021 and processed using the lidR open-source package. The right panel shows the canopy height model derived from the point cloud at a spatial grain of 1 m².

Radar Sensing

Radar is another type of active data collection using emitted microwaves by the system to determine the distance, angle, and physical properties of objects on the Earth's surface. An active radar sensor does not rely on sunlight as the radiation source and can therefore operate even at night and in cloudy conditions. Although radar methods themselves are not new and historically have been used for topographic mapping, radar data has played an increasingly prominent role in ecosystem dynamics research due to the emergence of open-access spaceborne radar systems, such as the ESA Sentinel-1 satellite, high spatial resolution of modern radar systems (<10 m), and the radar's ability to penetrate cloud cover.

Recent studies have tested the ability of radar to track vegetation phenology, forest disturbances caused by bark beetles and fires, as well as deforestation in the tropics, where the ability to see through clouds has significantly expanded monitoring capabilities. Depending on radar polarization, radar can penetrate vegetation and thus assess vegetation structure. Radar has also been applied to study wetlands and urban areas, thus its utility extends beyond more classical applications in forest ecosystems.

While additional research is needed, radar holds great promise for analyzing vegetation structure in space and time, regardless of cloud cover. Thus, future radar systems are specifically designed for global monitoring of biomass dynamics.

Fig.7. Example of seamless radar data set acquired over the European Alps from Sentinel-1. Red/green/blue colors show different radar signal polarizations, averaged for the summer of 2021.

Thermal Sensing:

Thermal cameras mounted on drones or satellites record infrared radiation, which is a reflection of the thermal radiation of objects on the Earth's surface. Thermal sensing allows researchers to study the thermal characteristics of ecosystems, including thermal anomalies, changes in the temperature of water resources, and heat distribution in urban areas.

Using thermal radiation to study processes related to the thermal balance in ecosystems. Detecting thermal anomalies and identifying sources of heat. Using for studying climate change and monitoring the thermal state of water resources.

Fig.8. – Rasters of noon surface temperatures in degrees Celsius and vegetation index NDVI for the seasons of 2011, May (a), September (b).

These land remote sensing methods have unique capabilities for obtaining data on the Earth's surface and its condition. Researchers can use this data to monitor ecosystems, detect changes in the environment, and make informed decisions in environmental protection.

Application of Remote Sensing Methods in Ecological Research

Mapping Forest Distribution. Application of Remote Sensing in Forest Management and Biodiversity Conservation.

Images and subsequent mapping are widely used in forest ecology and restoration plans. Additionally, researchers rely on remote sensor data to assess biophysical and biochemical properties, down to individual trees in the forest. Currently, biodiversity needs to be protected primarily to maintain the mechanisms of functioning of living nature in forests and ecosystems; maintain the ability to withstand environmental changes, as well as to discover and use new opportunities that can contribute to the development and ensure the survival of future generations. The evolution of remote sensing tools allows for the improvement of existing approaches and the development of new innovative approaches for better assessing the response of biodiversity to the management of natural ecosystems and their preservation.

Fig.9. Forest mapping based on images down to the tree level

Land Use Information

In ecology, researchers seek to understand the relationship between living organisms and their physical environment. To do this, they must study vegetation cover and its use. There are "thousands" of ways to do this, but one of the most convenient and reliable is remote sensors. Remote sensors create images necessary for analyzing how land is used, and advanced LULC technology provides detailed indexes of its composition and critical changes towards deterioration. Images also clearly show the land surface, which is traditionally mapped to analyze the environmental situation (including mapping land composition indexes) and allows understanding the scale of potential environmental damage.

Fig.10. Mapping of ecological land use changes

Overall, the pace of land degradation on Earth today is still higher than the pace of land restoration. Despite the potential for climate change mitigation through global tree restoration, there is an urgent need for action. Rehabilitating 12 million hectares of degraded land per year can help reduce the emissions gap by 25% by 2030.

Given the need for accelerated action on a global scale, land planners, policymakers, and other stakeholders need the best available data to understand, plan, and manage ecosystem restoration and accurately calculate carbon sequestration in these ecosystems.

Plant Chemistry and Moisture

When studying vegetation and plant moisture, various tools are used. The chemical composition of vegetation is a crucial indicator of their ecological condition. Multispectral instruments with high and coarse spatial resolution, which are remote sensors, work wonders in studying the chemical composition of vegetation and moisture. Botanists rely on this sensor (detector) as the primary source of ecological data. Based on the synthesis of various multispectral channels, information about the state of vegetation can be obtained.

Fig.11. Absorption bands of chlorophyll by plants

Fig.12. - Reflection spectra at different levels of chlorophyll content in the leaf

Carbon Inventory

Assessments of biomass stocks are required for carbon inventory, and remote sensing aids in obtaining the information or data needed for such assessments or for validating assessments made by other means. Biomass stock assessments are based on vegetation characteristics such as coverage and slope.

One of the most critical issues when using remote sensing data is the accuracy of assessments. Concern also extends to carbon stock assessments. Accuracy of 80–95% is achievable using high-resolution images (1-10m) in differentiating forested and non-forested areas. However, achieving such high accuracy in determining biomass and, consequently, carbon content is much more challenging – the most reliable carbon stock assessments are those based on UAVs and field measurements.

Carbon is a primary component of plant biomass and is invisible. Therefore, to estimate carbon stocks, it is necessary to rely on vegetation characteristics. These characteristics may include vegetation age, tree diameter, chlorophyll activity intensity, or biomass density.

Fig.13. Relationship between field-assessed aboveground biomass and vegetation indices, obtained with Sentinel-2, such as NDVI and MSAVI for each plot.

Fig.14. Carbon stock map of shrubs in the western desert region of Mongolia. Pixel area 100 m².

Average carbon stock is 5.1 kg/pixel in this region.

Fig.15. CO2 concentration highlighted in color.

Fig.16. Example spectral reflectance plot of sphagnum – carbon biomass.

Visible and near-infrared ranges correspond to wavelengths used by Landsat (blue 450–515 nm; green 525–600 nm; red 630–680 nm; NIR 845-885 nm). Peaks of vegetation fluorescence (690 and 740 nm) and water reflection trough (950–970 nm, used for WI) are added. The sample was taken from the Forsinard Flows RSPB reserve, and the reflectance was obtained in the laboratory using a Ger3700 spectrometer. It is worth noting that the reflectance of sphagnum-carbon strongly depends on water content, bleaching, and increased reflectance in all wavelengths as it dries.

Conclusion

From the perspective of a scientist and an ecological practitioner involved in remote sensing, ecological landscapes consist of pixels, sometimes concealing the complexity of the real world. Therefore, it is important for individuals primarily working with remote sensing data to also go out into the field and assess the complexity of ecological systems in field conditions.

Remote sensing allows for a better quantitative assessment of ecosystem heterogeneity (i.e., spatial and temporal variations in ecological processes) and, thus, significantly improves understanding of ecosystem resilience.

For example, continuous monitoring of vegetation indices in space and time can detect early warning signals, such as enhanced temporal autocorrelation, which may indicate a sudden change in ecosystem state. In this regard, the ability of remote sensing to track spatial patterns over time (e.g., forest fragmentation) or the composition and configuration of disturbance patches opens up new avenues for quantitative assessment of ecosystem resilience beyond traditional measures.

Remote sensing can also assist in directly quantifying ecosystem resilience, for example, by mapping the rate of its recovery after disturbances.

The contribution of remote sensing to accurate mapping of microclimates is significant, improving subsequent predictive models.

Considering these research directions and better education of ecologists in data and remote sensing analysis, it can be affirmatively stated that remote sensing has become an important tool for studying ecosystem states across the globe.