Source report sheet: Status_20260428
The main goal of this activity is to prepare selected instruments (PanSels) so that the studies performed in Activities 1 and 2 can be applied and tested. This preparation includes expedited calibration, specialized calibration—such as spectral stray light calibration using a tunable laser and cut-on filters at the LuftBlick laboratory—expedited quality assurance and control (QA/QC), and accelerated data processing..
The instruments list can be found in this google sheet.
25 Pandora instruments have been chosen as “PanSels”, of which 10 are Pandora-2S models. The selection was made based on geographical distribution, pollution level relevance, minimum 1 year of data availability, presence of co-located instruments (AERONET, Aerosol LIDAR, Brewer), and suitability for different work packages (1.4, 1.5, 1.6). The instruments list can be found in this google sheet.
| Total PanSels selected | 25 |
| Pandora-2S instruments | 10 |
| Co-located PanSels | 23 |
| Average processed days (p1-9) | 1186.2 |
| Average L0 data available (as of 28-Apr-2026) | 2119.1 |
| Average QA [%] | 62.8 |
| PanSels with QA > 90% | 9 |
| PanSels with QA > 70% | 13 |
The map illustrates the global distribution of these 25 Pandora sites, color-coded by the location type. The classification of location type is based on the environmental and urban conditions. Key criteria include: geographic diversity ensuring global coverage from arctic to tropical regions; environmental setting which includes remote, rural, semi-urban, and urban locations; and special atmospheric conditions such as high-altitude, coastal, mountain-surrounded, and polar sites.
Section 1b separates the overall co-location share from the exact co-located instrument combinations. AERONET is the most common combination (15), followed by the remaining combinations shown in the bar chart.
Overall, 92% (23 out of 25) of the PanSels are co-located with at least one high-quality reference instrument. This enhances the robustness of the network and improves validation purposes by enabling cross-comparisons and synergy.
The number of days or entries of raw data available (L0) together with the number of processed days for p1-8 and p1-9 for each PanSel instrument. A close match between brown and green bars in this figure indicates that most of the available data was also processed, thereby indicating the data throughput and processing pipeline efficiency.
Network and hardware maintenance checks are ideally scheduled once per week for all instruments in the PGN. However, due to operational constraints such as personnel availability, access issues, or communication downtime, maintaining this frequency across all PanSels is often difficult. More info
In contrast, QC checks are generally carried out less frequently. Within the broader PGN, QC assessments are performed approximately once per month. For the purposes of this project, the PanSels have been flagged and categorized by instrument type, allowing QOs to target them specifically. The planned frequency for QC checks on PanSels is twice per month, supporting enhanced calibration accuracy and early anomaly detection. At the time of reporting, 68.0% of the PanSels are operational, 12.0% are operational with issue, are on hold due to issue, and are out of operation. For QC checks, 64.0% have been reviewed within a 1-2 week time frame. For network operator checks, 16.0% were updated within a 1-2 week time frame. More info
Calibration of Spectrometer 1 (S1) has been successfully completed for 24 out of the 25 PanSels using the current operational version 1.8 of the processing software. This represents a 96.0% completion rate, with the remaining 1 instruments either partially calibrated or pending final validation. For S1 in version 1.9, 14 out of 25 instruments are fully calibrated, 4 fall in the 75-95% calibration coverage range, and the remaining 7 are below that range or still ongoing. Among the 10 sites with both S1 and S2 available, 5 have both spectrometers fully calibrated in v1.9. The average calibration coverage is 95.7% for S1 v1.8, 84.8% for S1 v1.9, 0.0% for S2 v1.8, and 82.1% for S2 v1.9.
As of the latest analysis, a significant portion of the PanSels exhibit delays in quality assurance (QA) updates. While the project plan targets every three months QA checks for each instrument and a full QA review is planned for PanSels without any QA, the actual percentage of quality-assured data across the 25 PanSels is currently highly variable, with 5 instruments showing QA coverage below 20%. The average QA coverage across the PanSels is 62.8%.
Time series of L2 data quality for the PanSels based on S1 (and S2) measurements. Grey bars indicate days with available L0 data, while blue bars show AERONET data availability. Colored markers represent the daily weighted mean WRMS from high-quality L2 observations. The WRMS is derived from H2O, NO2, O3, SO2, HCHO, and O2O2, with NO2 and O3 double-weighted. Percentage values denote the fraction of high-quality L2 data, and sites are ordered by their overall WRMS. Each subsection embeds the live PanSels multiplot and also provides a direct link.
https://reports.hetzner.pandonia-global-network.org/PanAMA/PanSels_Multiplot_S1.html
https://reports.hetzner.pandonia-global-network.org/PanAMA/PanSels_Multiplot_S2.html