Environmental monitoring spans a vast and growing toolkit of technologies used to track what is happening in the air, water, soil, and living ecosystems around us. From satellites orbiting hundreds of kilometers overhead to sensors strapped to the backs of elephant seals, the field has expanded well beyond the manual sampling campaigns that dominated for most of the twentieth century. What makes the current era distinctive is not any single breakthrough but the convergence of cheap sensors, satellite coverage, biological detection methods, and networked data systems that together allow researchers and regulators to watch environmental change in something close to real time.
Watching From Above
Satellite remote sensing remains the backbone of large-scale environmental monitoring. For ocean systems, decades of daily satellite-based sea surface temperature data now allow researchers to identify marine heatwaves and cold spells across the globe, though the identification of these extremes depends heavily on which dataset and algorithm a researcher chooses, making objective comparison difficult.1Journal of Geophysical Research: Oceans. Detection of Satellite Sea Surface Temperature Extremes: Low Frequency Variability and Climate Change That sensitivity to methodological choices is worth keeping in mind whenever you see headlines about record ocean temperatures: the trend is real, but the precise numbers shift depending on how the data are processed.
For ice and glaciers, space-borne interferometric synthetic aperture radar (InSAR) has been transformative. The technique measures surface motion of glaciers and ice sheets at wide spatial resolution, day and night, regardless of cloud cover.2Advances in Remote Sensing. A Review on Applications of Imaging Synthetic Aperture Radar with a Special Focus on Cryospheric Studies Tracking how quickly a glacier is sliding toward the sea used to require months of fieldwork in dangerous terrain. Now it can be done from orbit on a continuous basis.
Closer to the ground, drones carrying lidar and hyperspectral cameras are filling in what satellites miss. In a study of forests in the southwestern United States, fusing drone-based lidar with hyperspectral imagery produced about 88% overall accuracy in classifying tree species, outperforming either data type alone. The lidar data on its own yielded strong estimates of individual tree height and canopy cover, which is useful for tracking recovery after prescribed burns or thinning treatments.3Elsevier / ScienceDirect. UAV lidar and hyperspectral fusion for forest monitoring in the southwestern USA Where satellites give you the big picture, drones give you the fine grain: individual trees, small erosion channels, localized pollution plumes.
Monitoring the Air We Breathe
Industrial smokestacks have been monitored with optical sensors for years, but the instruments keep getting sharper. Non-dispersive infrared spectroscopy can now simultaneously track gases like carbon dioxide and carbon monoxide in stack emissions while correcting for the ways different gases interfere with each other’s measurement signals.4Atmospheric Measurement Techniques. Stack emission monitoring using non-dispersive infrared spectroscopy with an optimized nonlinear absorption cross interference correction algorithm For sulfur dioxide, broadband ultraviolet absorption sensors have reached detection limits as low as one part per million and have been tested directly in flue gas from coal-fired boilers, including in the presence of suspended particles.5Applied Optics. Broadband spectroscopic sensor for real-time monitoring of industrial SO2 emissions These instruments are compact and do not require calibration cells, which makes them practical for continuous field deployment rather than occasional spot checks.
Away from industrial stacks, a parallel revolution is happening with low-cost consumer-grade air quality sensors. In Accra, Ghana, researchers tested inexpensive AirGradient monitors against a reference-grade instrument and found the low-cost devices overestimated fine particulate matter concentrations by about 34%. After applying a calibration model, however, the error dropped dramatically, suggesting that cheap sensors can deliver useful data if properly corrected against a reliable reference.6ACS ES&T Air. Urban Air Quality Management at Low Cost Using Micro Air Sensors: A Case Study from Accra, Ghana This matters because many cities in the Global South have little or no regulatory air quality infrastructure. A network of calibrated low-cost sensors could provide neighborhoods with actionable pollution data for a fraction of what a traditional monitoring station costs.
Keeping Tabs on Water Quality
Traditional water quality testing involves someone collecting a bottle of water, driving it to a lab, and waiting for results. By the time you learn something is wrong, a contamination event may already be over. Internet of Things (IoT) platforms are changing that by placing sensor arrays directly in rivers, reservoirs, and treatment plants. These systems continuously measure parameters like pH, dissolved oxygen, total dissolved solids, and temperature, then transmit the data through wireless networks to cloud-based platforms for analysis.7Heliyon. IoT based real-time water quality monitoring system in water treatment plants (WTPs) Similar sensor-based systems have been proposed for rivers, replacing the old manual approach that was slow and labor-intensive.8Procedia Computer Science. IoT Based Real-time River Water Quality Monitoring System
The appeal is obvious, but the hardware has limitations. A study testing 19 different low-cost water quality sensors from four manufacturers found variable performance across parameters like pH, dissolved oxygen, turbidity, and oxidation-reduction potential. Some sensors tracked well against a reference multiparameter probe; others drifted or showed poor precision.9PubMed. Evaluation of low-cost sensors to integrate in a water quality monitor for real-time measurements The lesson is the same as with air sensors: low-cost does not automatically mean low-quality, but you need to validate rigorously and calibrate often.
The Biofouling Problem
If you leave any sensor submerged in natural water for more than a few weeks, organisms start colonizing its surface. Algae, bacteria, and invertebrates form biofilms that degrade measurement accuracy, sometimes rendering a sensor useless within a month. This biofouling problem is one of the biggest practical obstacles to long-term autonomous water monitoring. Current mitigation strategies range from mechanical wipers and coatings to chemical and biological approaches, each with trade-offs in cost, effectiveness, and environmental impact.10PubMed Central. Antibiofouling Coatings For Marine Sensors: Progress and Perspectives on Materials, Methods, Impacts, and Field Trial Studies
A promising new approach uses surface acoustic wave (SAW) devices to keep sensor windows clean. A small, inexpensive device operating at 16 MHz was able to reduce biofilm formation by adhesive diatom and planktonic algae strains by roughly 98% compared to untreated surfaces, without harming the surrounding aquatic organisms. The device can run for a couple of months without maintenance and consumes very little power.11PubMed Central. Exploring Surface Acoustic Waves (SAWs) for Water Quality Sensor’s Anti-Biofouling Application: A New Direction for Acoustic Waves If this scales, it could mean the difference between a monitoring station that fails after four weeks and one that delivers reliable data for an entire season.
Listening to Ecosystems
Not all monitoring involves measuring chemicals or temperatures. Passive acoustic monitoring (PAM) records the sounds of an environment and uses them to track which species are present, how active they are, and how communities change over time. A systematic review of terrestrial PAM covering hundreds of published studies found that the technique was initially dominated by bat research but has been diversifying rapidly across taxa and regions. Advances in sensor hardware and machine learning for automated call identification are pushing the method forward.12Methods in Ecology and Evolution. Emerging opportunities and challenges for passive acoustics in ecological assessment and monitoring The appeal of PAM is that a single recorder left in a forest or on a reef can capture far more species encounters than a human observer could, operating around the clock without disturbing wildlife.13BioScience. Terrestrial Passive Acoustic Monitoring: Review and Perspectives
Despite its growth in applied conservation, PAM’s potential contribution to fundamental ecology is less often explored. Researchers have argued the technique could illuminate basic ecological questions about community assembly, competition, and behavioral ecology that are hard to study through visual observation alone.14Functional Ecology. Passive acoustic monitoring provides a fresh perspective on fundamental ecological questions The bottleneck is no longer hardware; it is processing. A single recorder generates enormous volumes of audio, and sorting signal from noise still requires substantial computational effort.
DNA in a Bottle of Water
Every organism sheds traces of its DNA into the environment through skin cells, mucus, feces, and decomposition. Environmental DNA (eDNA) metabarcoding captures those traces from water samples and sequences them to identify which species are present. The technique has shown particular promise for early detection of invasive species, since it can pick up genetic signatures from organisms at low population densities before they are noticed by traditional surveys.15PubMed Central. Integrating invasive species risk assessment into environmental DNA metabarcoding reference libraries
A targeted metabarcoding approach designed to detect 69 invasive fishes and aquatic invertebrates achieved detection of about 98% of species in spiked DNA samples, with read counts positively related to the quantity of DNA present.16Environmental DNA. Development and validation of targeted environmental DNA (eDNA) metabarcoding for early detection of 69 invasive fishes and aquatic invertebrates In a real-world application at 16 major Canadian ports, metabarcoding identified 379 zooplankton species, including 24 non-indigenous species, 11 of which were found in locations where they had not been previously reported.17Diversity and Distributions. Early detection of aquatic invaders using metabarcoding reveals a high number of non‐indigenous species in Canadian ports That kind of surveillance would be essentially impossible with nets and microscopes alone.
The method has limitations. Comprehensive reference databases of known species’ genetic sequences are still incomplete, which means eDNA can miss species that are not well represented in existing libraries. And closely related species can be hard to tell apart with some barcode markers. The technology works best as a complement to, not a replacement for, traditional biological surveys.
Living Sensors
Some of the oldest forms of environmental monitoring use living organisms as indicators. Lichens, which absorb nutrients and pollutants directly from the air, have long served as biomonitors for atmospheric contamination. In the eastern United States, specific lichen species are recommended as element indicators of local air pollution loads, accumulating nitrogen, sulfur, and metals in proportion to what is in the air.18Plant and Fungal Systematics. Lichen species as element bioindicators for air pollution in the eastern United States of America Around Salda Lake in Turkey, analysis of the lichen Xanthoria parietina revealed significantly elevated levels of nickel, iron, aluminum, copper, and zinc compared to reference values, and spatial mapping of those metals provided pollution maps for the region.19Air Quality, Atmosphere & Health. Biomonitoring of atmospheric heavy metals in native lichen Xanthoria parietina around Salda Lake (Burdur – Turkey), a special environmental protection area Lichens will never replace electronic instruments for real-time alerts, but they offer something instruments cannot: a time-integrated record of what an organism actually experienced, which can be more biologically meaningful than an instantaneous sensor reading.
At the other end of the size spectrum, marine animals themselves are becoming monitoring platforms. Hundreds of seals in the Southern Ocean have been equipped with conductivity-temperature-depth sensors, collecting and transmitting an average of 500 temperature-salinity-depth profiles per animal per year. When instruments are recovered (roughly 30% of those deployed annually), they can yield up to a thousand profiles per month.20Frontiers in Marine Science. Animal Borne Ocean Sensors – AniBOS – An Essential Component of the Global Ocean Observing System These instrumented animals efficiently fill critical observational gaps in polar regions where traditional oceanographic platforms are scarce due to ice cover and logistical costs.21Geophysical Research Letters. Estimates of the Southern Ocean general circulation improved by animal‐borne instruments The data have meaningfully improved estimates of Southern Ocean circulation patterns. A seal goes where a research vessel cannot, dives repeatedly to depth, and does it all while going about its normal life.
Wastewater as a Community Health Mirror
One of the monitoring applications that gained widest public attention during the COVID-19 pandemic was wastewater surveillance. Since a large fraction of infected individuals shed viral RNA in their feces, sewage samples can serve as a population-level indicator of disease prevalence. A meta-analysis of COVID-19 studies found that about 48% of patients had detectable SARS-CoV-2 RNA in their feces, which made wastewater a surprisingly sensitive tracking tool for community transmission trends.22Frontiers in Public Health. A narrative review of wastewater surveillance: pathogens of concern, applications, detection methods, and challenges The approach has since been extended beyond COVID. Sampling from wastewater treatment plants in Australia detected 13 different respiratory viruses, including bocavirus, parechovirus, and two rhinovirus types in all samples tested.
Wastewater surveillance sidesteps a core limitation of clinical testing: it captures signals from people who never get tested, including asymptomatic carriers. A single sewage sample reflects thousands of individuals at once, which makes it a powerful early-warning system for outbreaks. Many cities that built wastewater monitoring infrastructure during the pandemic are now expanding it to track influenza, RSV, and other pathogens year-round.
Citizen Science and the Quality Question
Governments and research institutions cannot maintain enough monitoring stations to cover every stream, lake, or neighborhood. Volunteer monitoring programs fill some of those gaps, but the perennial concern is data quality. A large-scale comparison between citizen science water quality data and professional measurements across an entire statewide program found 81% overall agreement for conductivity, pH, and dissolved oxygen. Conductivity showed the best match at 86%, while dissolved oxygen was lowest at 77%. After adjusting for known biases, overall agreement climbed to 91%.23PLoS ONE. Accuracy of long-term volunteer water monitoring data: A multiscale analysis from a statewide citizen science program The results suggest that volunteer-collected data, gathered under routine protocols rather than special experimental conditions, can be comparable to professional data for many purposes.
The practical implication is significant. Many regulatory agencies are reluctant to use citizen science data for decision-making because of perceived quality concerns. Studies like this suggest the concern is overstated, at least for well-structured programs with standard protocols and basic training. The bigger barrier may be institutional culture rather than data reliability.
The Data Interoperability Challenge
Collecting environmental data is increasingly the easy part. Making it usable across programs and agencies is harder. An effort to integrate data from four large-scale, long-term US aquatic monitoring programs found that differences in data governance resulted in considerable effort just to access and reuse the original datasets. Requirements, guidance, and resources for data publishing were inconsistent across agencies, producing various formats stored in different locations that were difficult to find, access, or combine.24PubMed. Sharing FAIR monitoring program data improves discoverability and reuse
The push toward FAIR data principles (findable, accessible, interoperable, reusable) is an attempt to fix this. But standardizing data formats after decades of independent program operation is unglamorous, slow work. Environmental monitoring generates staggering volumes of information: satellite imagery, continuous sensor streams, acoustic recordings, genetic sequences, volunteer observations. The value of all that data increases enormously when it can be cross-referenced and combined, but right now much of it sits in incompatible silos. A wireless sensor network in a forest might record temperature and humidity at high resolution, but if its data cannot be easily merged with satellite imagery or soil carbon measurements from the same site, the integrated picture that researchers need stays incomplete.
Ground-Level Networks and Soil
Wireless sensor networks deployed inside forest canopies can capture microclimatic variation that satellites miss entirely. Nodes recording temperature and humidity at high resolution (accuracy of ±0.3 °C for temperature and ±3% for relative humidity) reveal gradients between sunlit openings and shaded understory that matter for species distribution and fire risk.25Scientific Reports. Evaluation of Microclimatic Detection by a Wireless Sensor Network in Forest Ecosystems These fine-grained ground measurements also serve as calibration points for broader satellite and drone datasets.
Soil monitoring is another area where simple, inexpensive approaches are proving surprisingly effective. For estimating soil organic carbon, which is a key indicator of soil health and carbon storage capacity, researchers have found that a standard digital camera measuring soil color can produce predictions with accuracy comparable to laboratory spectroscopy.26Geoderma. Spectroscopic measurements and imaging of soil colour for field scale estimation of soil organic carbon Darker soils generally indicate higher organic carbon content, and a camera can capture that variation across a field much faster than point-by-point lab analysis. It is a reminder that environmental monitoring does not always require expensive instrumentation; sometimes the limiting factor is clever use of what already exists.