Environmental conditions can interfere with satellite methane measurements

Local environmental factors, from cloud-cover to rough terrain, can limit satellites’ ability to detect methane, underscoring the need for alternative measurements in certain areas.

While methane is invisible to the human eye, it absorbs infrared radiation and can be observed with specialised cameras.  Such instruments are now mounted on satellites where they observe the sunlight that passes through the atmosphere, reflects off the earth’s surface, and travels again towards space, as illustrated in the following figure. They are tuned to search this light for specific absorption patterns that serve as a fingerprint of methane in the atmosphere. Specialised algorithms are then used to consider many other factors that could influence the amount of radiation that reaches the satellite and infer the amount of methane that was present in the light’s path. 

Several environmental factors can stop sunlight from reaching the satellite sensor and these  may make methane detection more challenging or impossible at times; other factors, like strong winds, can disperse and dilute methane in the atmosphere, hindering methane detection – for example, leading to higher detection limits.

The impacts of environmental conditions are satellite-specific

Satellites are not all equally affected by diverse environmental conditions. Τhe exact way that these conditions will affect satellites’ ability to measure and quantify methane emissions will vary depending on each instrument’s specific characteristics. For example, instruments built to detect changes in methane concentration over large areas will be greatly affected even by a few clouds in the area they are studying; at the same time, a high-resolution satellite, trying to detect emissions from a facility in the same area, might be able to see between broken clouds and observe the facility.

Researchers have been long studying the ways that observation conditions will affect different methane-sensing satellite instruments. For example, recent research has highlighted the dramatic impact that a sensor’s pixel size will have on its ability to observe methane in the tropics. Other researchers have studied in detail the factors that limit the ability of a specific instrument (TROPOMI) to observe methane around the world. 

This report summarises such studies and highlights the broader patterns that affect methane satellite observations in one way or the other by mapping the regions where satellites are most and least affected by these environmental issues, without focusing on the characteristics of specific instruments. 

Consequently, we label observation conditions using the generic categories of “favourable”, “moderate” and “difficult.”. In “favourable” conditions — cloud-free skies and relatively flat bright surfaces — satellites are expected to perform their best. When areas are marked as  having “moderate” conditions, methane sources might be frequently covered by clouds and the complex terrain might stop satellites detecting some plumes they would otherwise easily spot. In “difficult” conditions, detecting methane might be impossible for some sensors or happen less frequently and/or for very large plumes.  

The exact impact of these environmental conditions on specific satellites should be studied separately for each use case and location.

Environmental conditions and their impact

This report considers the following environmental factors and their impact on the ability of satellites to effectively detect methane.

Cloud cover

Methane satellites cannot see through clouds and this makes cloud cover the most important obstacle for detecting methane from space. Since they are much brighter than the earth’s surface, even a small cloud partly covering the satellite pixel will interfere with the measurements and stop the satellite from detecting methane. In some regions with persistent cloud cover, like the tropics, this can hinder  satellite  methane observations  year round.

Satellite characteristics, like their spatial resolution, will greatly influence how clouds impact their ability to monitor methane. For example, a high-resolution satellite, built to monitor methane from individual facilities, might be able to detect methane between broken clouds, although challenges remain with these observations. In contrast, a satellite built to quantify methane in wider regions, if faced with the same cloud cover, might never find a completely cloud-free view that is needed for its measurements. 

Terrain roughness

Some satellites find it challenging to estimate methane in regions with mountainous terrain. The uneven ground in such regions will create shadows in part of the observed scene, making it hard to interpret the amount of light that reaches the satellite. As light reflects off different parts of the uneven terrain, it will reach the satellite having passed through different parts of the atmosphere, further complicating the interpretation of satellite signals. In combination, these factors make rough terrain a year-round obstacle for methane sensing. As the map shows, this can affect large parts of the world, including central Asia, western South America, mid-Western North America and the western Balkans. 

As with clouds, the way this parameter affects different satellite sensors will greatly depend on its sensor characteristics. A high-resolution sensor might be able to quantify methane over a small flat area around a facility of interest while, in contrast, an area-monitoring satellite might find it impossible to estimate methane emissions over the wider mountainous region.

Sun elevation

Most methane-detecting satellites that are used to study near-surface methane concentrations require bright sunlight to operate effectively, however,  sufficient light is not always available. During winter at high latitude, for example, the sun does not rise high above the horizon, leaving large parts of the earth poorly lit for several months at that time. As a result, data coverage can be significantly reduced during winter for some sensors. Parts of Russia, Canada, and Argentina are strongly affected by this parameter.

The exact impact of low light conditions on each methane-sensing instrument will depend on its design. Instruments that are optimised to operate in low-light conditions might be less affected, but will be negatively impacted in any case. The satellite orbit, which defines the local time that the satellite will orbit over a location, will also affect the impact of this parameter on the specific instrument.

Surface brightness

Dark ground surfaces, like forests, will not reflect enough light back towards the satellite sensor, possibly leading to higher uncertainty.  As with low sun elevation, this might make it difficult to detect methane absorption. Snow-covered areas also appear mostly dark in the shortwave infrared wavelengths used to detect methane. This effect can be easily seen over tropical forests and snow-covered regions at high-latitudes. In contrast, arid regions will reflect abundant light and be favourable for methane sensing.

Inland and ocean water surfaces are excluded in our study, because, in most cases, they are too dark to be used for methane retrievals. Note that new techniques are being developed to adjust the viewing angle of the sensor to enhance reflected sunlight off the water (sun glint). This technique can offer significant insight e.g. for offshore oil and gas production, but is not part of this report. For some instruments, a surface that is too bright also poses a challenge, as the large amount of light could blind the sensor; since this is uncommon, the effect will not be accounted for in our study.

Wind speed

Strong winds can quickly disperse methane far from its source, reducing local concentrations and thus making it harder for a satellite to distinguish the plume from background methane concentration. While a certain amount of wind is needed for plume formation and thus plume quantification, in general, higher wind speeds are associated with higher detection thresholds.  

 In regions with persistent strong winds, like in southern Argentina, satellites will only be able to detect larger plumes compared with regions with milder wind conditions.

Wind conditions will have a greater impact on high-resolution plume-imaging satellites, as methane plumes will be quickly dispersed by strong winds. Coarse resolution satellites, aiming to map methane over wider regions, will be less affected by wind conditions.

Our methodology

These five environmental factors, and their change over the course of a year, have been studied to explore the favourability of locations for methane observations.  First, the effect of each environmental factor for each month is studied individually. Specifically, we apply thresholds to each of the five factors to define if the conditions at that particular time and location are ‘favourable’, ‘moderate’, or ‘difficult’ for observations (see the Appendix for a detailed description of the thresholds and input datasets used). Secondly, we calculate the combined impact of all parameters for each month. If any input falls into the ‘difficult’ category, the month will be classified as difficult. Similarly, if any input is categorised as ‘moderate’ but none as ‘difficult’,  the final score will be ‘moderate’. To achieve a ‘favourable’ score, all inputs must be in the ‘favourable’ category. The result is a monthly classification for each location worldwide.

As noted before, the assigned categories should be used only as rough indicators of areas where satellite observations of methane may be less available . Our analysis is conservative in the sense that we are identifying areas and times of the year where at least some satellite sensors are expected to be less able, or unable, to monitor methane. These categories do not, for example, capture how environmental conditions differently impact instruments designed for monitoring facility-scale emissions versus those intended to measure emissions over larger areas. In areas labelled as “moderate” or “difficult” to monitor, users need to be attentive, as some satellites that otherwise may be very useful for their use case may not perform as expected.

Next, we investigate the impact of these environmental factors on coal, oil, and gas extraction infrastructure, responsible for a substantial portion of anthropogenic methane emissions.