To improve diagnosis of Amazonia's carbon cycle, we present a 2 year data (2019 to 2020), complementary to our previosly 9 years (2010 to 2018) observations of lower troposphere CO2 concentrations performed regularly at four aircraft vertical profiling sites spread over the Brazilian Amazonia (Gatti et al., 2021). The four sites from the CARBAM project at Amazonia: SAN (2.86S 54.95W); ALF (8.80S 56.75W); RBA (9.38S 67.62W) and TEF (3.39S 65.6W), started in 2013. The sampling period was typically twice per month (Gatti et al., 2021; Gatti et al., 2014; Basso et al., 2016; Miller et al., 2007; d'Amelio et al., 2009; Domingues et al., 2020). From 2019 to 2020, a total of 141 vertical profiles from 4420 m to 300 m asl. was performed at the 4 sites. The vertical profiles were usually taken between 12:00 and 13:00 local time. Air is sampled by semi-automatic filling of 0.7 L boro-silicate flasks inside purpose-built suitcases (PFP - Programmable Flask Package) (Tans et al., 1996); there are two versions, one with 17 flasks at SAN, and another with 12 flasks at TEF, ALF and RBA. This suitcase is connected to a compressor package (PCP –Portable Compressor Package), containing batteries and 2 compressors, which is connected to an air inlet on the outside of the aircraft at wing or window, depending on the aircraft model. Once a PFP (i.e. one vertical profile) has been filled with air the PFP is transported to the INPE/ LaGEE(Instituto Nacional de Pesquisas Espaciais/Greenhouse Gases Laboratory), in Sao Jose dos Campos, Sao Paulo state, Brazil. This laboratory is a replica of the NOAA/ESRL/GMD trace gas analysis system at Boulder, Colorado, USA, and was constructed in 2003 and sent to IPEN where started the analysis in 2004.Air samples were analysed with a non-dispersive infrared (NDIR) analyser for CO2. To ensure the accuracy, we construct a calibration curve every 2 samples. The calibration curve constructed with 3-standards concentrations, produced by NOAA/ESRL/GMD. The "High" (10 ppm higher than medium), "medium" (similar to mean CO2 concentration founded in Amazonia), and "Low" (10 ppm lower than medium). We have an intercomparison program with NOAA at Natal site (5S, 35W, located at Brazilian northeast coast) where the comparison IPEN/INPE-NOAA was -0.05 ± 0.38ppm. The precision is analysed based on CO2mole fraction from "target tanks" (calibrated CO2 in air in high pressure cylinders treated as unknowns by NOAA) and demonstrated long-term repeatability of 0.03ppm and a difference between measured and calibrated values of 0.03 ppm. Additional information can be shared from the LaGEE/INPE group as temperature, precipitation, and others parameters used by the group for the Nature paper entitled "Decrease in Amazonia carbon uptake linked to trends in deforestation and climate" (Gatti et al, 2021).

Please contact the coordinator Luciana V. Gatti mailto:Luciana.gatti@inpe.br or mailto:lvgatti@gmail.comVertical profiles of CO2 and CO performed from 2019 to 2020 at the 4 sites in Brazilian Amazonia (SAN - 2.86S, 54.95W; ALF - 8.80S, 56.75W; RBA - 9.38S, 67.62W and TEF - 3.39S, 65.60W): 8 files.Background concentrations of CO2 and CO and time trajectories from 2010 to 2020 at 5 sites in Brazilian Amazonia (SAN - 2.86S, 54.95W; ALF - 8.80S, 56.75W; RBA - 9.38S, 67.62W; TEF - 3.39S, 65.60W and TAB - 5.74S, 69.90W), calculated using BKG Method (Domingues et al., 2020): 10 files.Shapefiles of the influence area (.tfw, .png, .tif and .txt files) for all sample sites in Brazilian Amazonia: 1 zip file.Total_C_Flux_uncertainty_Calculation.zip contains the scripts (R Language) and files necessary to calculate the Total C Flux uncertainty in each region of Amazon (1 zip file containing 14 files).