Blog

Crossing the ITCZ brought a stretch of turbulence. Over Europe, the route followed the Spanish coastline, passing the harbor of Barcelona — which was also a target of ESA’s AB2CD campaign in October 2025 with the French ATR42 — before continuing over the Po Valley in Italy, an area previously surveyed during CoMet 1.0. Combining measurements from different campaigns and locations helps to consolidate the results. HALO touched down at Oberpfaffenhofen in the early evening, bringing the CoMet 3.0 Tropics campaign to a close after 18 flights.

Flight N°17: Transfer to Fortaleza

Flight no, seventeen was not a scientific flight. As there was no Brazilian observer on board, the crew did not have clearance to take measurements on the way from Cuiabá to Fortaleza. Therefore, HALO flew the roughly three-hour transfer on normal commercial air routes, before the final flight home was due.

The remote-sensing leg started under clear skies, with the first leg entirely cloud-free and the second, south of the Amazon, being only lightly affected by cumulus. Conditions worsened later in the pattern, with clouds and cloud shadows affecting a large share of the MAMAP2D data on the final remote-sensing legs. Two dropsondes were released during the flight to profile the atmosphere.

During the subsequent in-situ leg, which involved flying low over hilly terrain, several fire plumes were encountered — one of which produced a clear signal across CO, CH₄, CO₂ and N₂O simultaneously. Further fires were tracked and measured as the aircraft continued its zigzag pattern along and across the Amazon River. A dedicated racetrack pattern concluded the in-situ part of the flight. Comparing the boundary layer with the free troposphere revealed elevated CO background levels, well above the roughly 60–80 ppb typical of the wet-season Amazon boundary layer. HALO then headed back to Cuiabá after a flight that once again combined pristine-forest carbon-cycle measurements with some productive, albeit opportunistic, fire chasing.

Mostly clear skies allowed the remote-sensing pattern to cover almost the entire target area. The following in-situ leg set a campaign record, with CH₄ mixing ratios above 2,900 ppb. High concentrations, however, do not necessarily mean a strong source. Low surface winds may have let CH₄ accumulate near the ground, and only further analysis will show how large the emissions really are. After the in-situ measurements, HALO overflew the Tefé flux tower at two altitudes. On the way back to Cuiabá, a well-timed cloud gap made it possible to survey an oil and gas facility as well. All in all, another data-rich flight to a new target region.

Hazy conditions dominated much of the transfer out. A fire that had been burning for at least six days in the southern Pantanal, covering some 75 km², was detected by the remote-sensing instruments. After descending into a spiral for the in-situ leg, the crew noticed the smell of smoke in the cabin and measured a CO enhancement of around 250 ppb upon entering the boundary layer. Five dropsondes were released during the flight to profile the atmosphere.

After passing south of a larger fire and spotting its plume, HALO turned back to fly directly through it. The plume had risen higher than estimated, but the instruments still picked up a sharp CO signal and MAMAP2D made clear K-line detections. Further along the route, the landscape became greener, with more lakes, and the river came back into view. HALO then made a short detour to fly over the AERIS ground station, which colleagues from DLR and UNEMAT had set up with meteorological and CH₄ instruments. A large feedlot was also passed along the way. As on earlier flights, CH₄ concentrations were concentrated in the same areas, reaching up to around 2,070 ppb. The flask samples collected during the flight captured both wetland and fire-related signals.

Hazy conditions in the morning raised the possibility of nearby fires, though none could be confirmed at the time. Cloud cover over the target area gradually dissipated as the flight progressed, so the remote-sensing pattern went ahead largely as planned. The ATTO racetrack pattern was completed at several altitudes, with dropsondes released both along the transfer and directly over the tower. Small to moderate fires were spotted in the area later in the flight, though none of their plumes crossed the aircraft's track. With time running short toward the end of the in-situ part, some of the planned final profiles had to be dropped. HALO still passed ATTO once more before heading back to Cuiabá, continuing remote-sensing measurements and releasing further dropsondes along the way.

Fires in the area pushed CO levels well above expectations during the transfer out, though the smoke plume itself stayed clear of the flight track. Conditions over the delta were far better than forecast. Clouds were patchy and mostly limited to land, so the remote-sensing pattern captured good data over the water. Over parts of the forest, however, clouds still reduced the coverage. Five dropsondes were released along the way, each returning a clean vertical profile of the atmosphere. Only one planned release had to be skipped.

The following in-situ leg showed that the concentrations of CH₄, CO, and CO₂ all increased  with increasing height above the water. CH₄ was distinctly elevated over wetlands threaded with small streams and over grasslands grazed by cattle near watering holes. Some of the CH₄ over open water appeared to have been carried in from nearby, densely vegetated islands. On the way back to Cuiabá, the crew also spotted a substantial fire near the intersection of two regional highways, concluding a long, and data-rich flight over both natural and human-caused CH₄ sources across the delta.

On arrival, the area was found to be almost entirely overcast, which was worse than forecast. However, the remote-sensing pattern went ahead anyway, and, fortunately, the clouds gradually broke up, allowing good datato be collected , with a lot of potential for cloud slicing studies for CHARM-F alongside several dropsonde releases. En route, the crew crossed several fire plumes, including one near El Perú that was not visible on the Copernicus fire watch. MAMAP2D detected this fire via K-lines, and MIRO recorded a CO peak of about 1,010 ppb. The JIG-JAS team also collected a flask sample directly from a small agricultural grass fire.

A planned missed approach at Porto Velho let HALO overfly the ground-station FTIR precisely as intended. The final target, a feedlot in Mato Grosso, was flown over five times. A strong signal was detected at all altitudes, which suggests good mixing within the boundary layer. The flight adds valuable time-series CH₄ data over the Bolivian wetlands.

Despite increasing convection, haze, low clouds and even some rain, the team completed all three planned dropsonde releases and all main measurement activities, while also observing several fires along the way. The flight additionally yielded useful data on the atmospheric boundary layer, supporting the interpretation of the trace-gas observations. All of this contributes to CoMet 3.0 Tropics' broader goal of quantifying CH₄ emissions from tropical ecosystems.

Flight No. 09: Back to the Pantanal wetlands in search of methane hotspots

Building on findings from our first Pantanal flight, HALO returned to the region on 16 August with two objectives: quantifying CH₄ emissions across the wetland and capturing data on biomass burning as a target of opportunity.

Despite a delayed take-off due to misunderstandings with ATC, the remote-sensing legs ran smoothly under clear skies. After spiraling down to the in-situ leg, the aircraft followed the Paraguay River valley north, where the team flew three experimental racetrack climb/descent maneuvers to determine the planetary boundary layer height. Signals over Bolivia were modest, but a 20ppb enhancement appeared while crossing back into Brazil, and CH₄ reached up to 2,400 ppb over the main river valley during the second racetrack pattern.

On the climb back to Cuiabá, an active fire was confirmed visually and picked up by the remote-sensing instrument as well. Despite early ATC communication hurdles, all scientific objectives were met, the racetrack maneuvers proved highly effective, and five dropsondes were released successfully.

Cuiabá serves as the campaign's central base for monitoring greenhouse gas emissions across the Pantanal, Cerrado, and Amazon biomes. Over six weeks, HALO is flying 14 research flights from its base, covering Brazilian and neighboring Bolivian airspace to measure atmospheric CO₂ and CH₄. Around 120 researchers and professionals are involved — 60 from Brazil and 60 from Germany — backed by roughly 10 million euros in European funding, including support from the accompanying ESA Earth Observation project BRACCIOLE, which studies CH₄, CO₂, N₂O, fire and vegetation photosynthetic activity and links the earlier CarbonARA campaign to CoMet 3.0 Tropics.

International project coordinator Andreas Fix explained why Mato Grosso matters so much for this research: the region sits at the crossroads of several ecosystems and emission sources, both natural, such as wetlands, and human-driven, such as agriculture. National coordinator Dirceu Luís Herdies (INPE) emphasized how the resulting data can refine climate models and support strategies for extreme events such as heatwaves and floods, gathered with instruments not otherwise available on the market.

Guests included Rodrigo Bruno Zanin (Adjunct Secretary of Science and Technology of Mato Grosso), Marcos de Sá (President of FAPEMAT) and Sérgio Roberto de Paulo (International Office, UFMT), along with colleagues from partner universities, research institutions and state secretariats. TV Gazeta covered the visit, bringing the campaign to a wider audience across Mato Grosso.

Thank you to everyone who joined us in Cuiabá, and to the press team for helping tell this story!

Real-time adjustments were needed due to airspace restrictions, but the instruments still captured a CH₄ enhancement of 140 ppb at the bottom of the spiral near Porto Velho, and the remote-sensing pattern over the Llanos de Moxos concluded with a missed approach at Trinidad, where the in-situ legs began.

Flat terrain, sparse tree cover, and virtually no infrastructure characterized a barren landscape. Across this landscape, substantial CH₄ enhancements were measured throughout the region. Along the way, HALO also passed through several fire plumes, which added useful CO₂ and CO data to the CH₄ measurements. Despite the early restrictions near Porto Velho, almost cloud-free skies over the Llanos de Moxos made this a highly productive flight.

Conditions turned out better than the forecast of low clouds, cirrus and possible deep convection had suggested, though cloud cover still made the remote-sensing leg over the ground stations challenging. Over the forest itself, skies were largely clear, which favored the CH₄/CO₂ column and SIF measurements. Six dropsondes were released — twice as much as originally planned — with all data received on the ground. After descending to in-situ altitudes, HALO flew a missed approach at Santarém airport, followed by an in-situ pattern that included overflights of the two EM27 ground stationsoperated by UFOPA.

The in-situ legs revealed a striking north–south CH₄ gradient, clear CO₂ eddies over the secondary forest, and a plume of elevated CO and CH₄. After a low loop over the LBA K67 tower, a spiral climb over the CUP tower rounded off the flight. An unstable CO₂ channel forced the CHARM-F lidar laser to be switched off at times. Nevertheless, HALO landed back in Cuiabá on time with a rich dataset. A second flight to the region, ideally also covering the primary forest, was already under consideration.

The night flight allowed the CHARM-F lidar's high-altitude remote-sensing patterns over ATTO and Balbina to be compared directly with in-situ data gathered later at sunrise, once the aircraft had spiraled down. At Balbina, this coordination extended to INPA and MPI-BGC, who carried out complementary lake measurements with support from UNEP's IMEO. At ATTO, the Large-Scale Biosphere-Atmosphere Program (LBA) kept all ground instruments running at the tall tower for the duration of the flight. HALO flew spiral profiles over ATTO while radiosondes were released for later comparison, along the Uatumã river, and over the Balbina reservoir the in-situ instruments recorded clear CH₄ and CO₂ enhancements. Balbina is suspected to be a significant methane source, and the data collected here will help pin down the processes behind it, despite some reduced visibility from haze along the way.

On 4 August, HALO took off for the campaign's fifth flight, heading to the Araguaia floodplain around the Bananal Island in the Cerrado region. The aim was to capture atmospheric profiles and greenhouse gas distributions across the river valleys, plus a few fire plumes for K-line measurements. In satellite data high methane signals show up in this area.

Cloud cover on the first leg soon gave way too much better remote-sensing conditions. Overpasses of active fires showed clear CH₄ and CO₂ enhancements alongside the river-valley, and six dropsondes were released to record temperature, humidity and pressure profiles. A successful flight all around, both scientifically and operationally.

The operational phase of CoMet 3.0 Tropics from Cuiabá got underway on 31 July, with HALO's first research flight targeting greenhouse gas emissions over the Pantanal in Brazil and Bolivia across multiple remote-sensing and in-situ legs.

Both remote-sensing and in-situ instruments picked up strong CH₄ enhancements over key geographical features along the river valleys, offering valuable insight into natural wetland emissions versus anthropogenic sources. Active fires along the route were also targeted directly, with multiple passes through smoke plumes at both remote-sensing and in-situ altitudes capturing further CH₄ enhancements.

Eight dropsondes were released over the course of the flight, and low-altitude sawtooth maneuvers mapped the boundary layer height before a scenic pass along the Brazil–Bolivia border river. HALO landed back in Cuiabá after eight hours in the air over the Pantanal. It was HALO's first breath of Bolivian air.

Flight N°03: Official start of CoMet 3.0 Tropics — from Fortaleza to Cuiabá

The flight from Fortaleza to Cuiabá completed HALO's journey to its campaign base, this time with a dedicated measurement program. Along the way, the aircraft crossed the Cerrado biome and the Serra da Mesa hydropower dam, providing a first test of the instruments over complex emission sources. Moreover, CH₄ and ammonia plumes were detected over a massive feedlot housing roughly 70,000 cattle, despite challenging wind conditions.

These baseline checks over the dam and the feedlot rounded off the transfer, ensuring that the aircraft and the instruments – which are now based in Cuiabá – were fully prepared for the forthcoming research flights.

Over southern France and Spain, particularly near Madrid, the flight tracked over plumes of smoke from the massive fires that plunged Southern Europe into fear in the summer of 2026. Leaving Europe, the long transatlantic leg from Seville to Fortaleza then offered an ideal opportunity for instrument calibration over water, with roll maneuvers used to evaluate back-reflection signals against the wind-roughened Atlantic Sea surface.

From wildfire plumes over Europe to ocean-glint calibrations over the open Atlantic, this first transfer flight provided extensive calibration data and initial insights into the atmosphere even before the actual campaign began.

More cloud cover than forecast complicated the USCB pattern, with parts of the downwind leg past the ArcelorMittal steel plant obscured. A missed approach at Kraków Airport still went ahead as planned, followed by an in-situ leg downwind of the shafts to sample the emission plume close to the ground. With precipitation and dense cloud over East Germany, the crew decided to skip the planned Jänschwalde power plant pattern and repeat the USCB pattern a second time instead, gathering an additional dataset over the coal mine ventilation shafts. On the way back, HALO still managed a fly-by past Lippendorf power plant under partial cloud cover, before finishing with a low pass over the Hohenpeißenberg ICOS station — a useful early test of the instruments and measurement strategy ahead of the campaign to come.

Successful EMI flight

An important step in preparing the CoMet 3.0 Tropics campaign was completed on schedule: the required EMI (electromagnetic interference) flight was carried out successfully, clearing HALO's instrument payload for the campaign ahead. A special thanks goes to everyone involved.

Kick-off workshop at USP, São Paulo

Long before HALO left the ground, the campaign took shape at a two-day workshop hosted by the Universidade de São Paulo (USP) — a major step forward for European–Brazilian science cooperation. Researchers from the DLR Institut für Physik der Atmosphäre, the University of Bremen, MPI-BGC, ESA, IPEN, INPE, CENSIPAM, and several leading Brazilian universities came together to share the latest findings on greenhouse gas emissions, advance flight planning for HALO, and build the scientific foundation for the campaign — spanning active and passive remote sensing, in-situ instruments, joint modeling strategies and outreach planning.

Alongside the science, preparing the campaign also meant securing the necessary permits for Brazilian airspace, a process carried out together with INPE and the German Embassy, and applying for the scientific license required from the Brazilian National Council for Scientific and Technological Development (CNPq). CoMet 3.0 Tropics is about more than the measurements themselves. It aims to build a lasting partnership between Europe and Brazil to better understand greenhouse gas fluxes and their drivers in tropical ecosystems such as the Amazon, the Pantanal and the Cerrado.