Hello to the research community!
When conducting 100% offline (no Wi-Fi) air quality monitoring campaigns outdoors, the microSD card goes from being a simple hardware component to becoming the sole safeguard for our primary data.
After reviewing the official PurpleAir documentation and cross-referencing it with Quality Assurance and Quality Control (QA/QC) principles in scientific instrumentation, I have compiled this summary of best practices for local storage management. I am sharing it to help safeguard the integrity of time series data in future projects:
Capacity and Native Format (FAT32): It is highly recommended to stay within the 32 GB limit. Although it is possible to force 64 GB or larger cards to FAT32 using third-party software (like Rufus or GUIFormat), 32 GB cards come with this format out of the box, minimizing compatibility issues. Furthermore, 32 GB is more than enough to store years’ worth of .csv files.
The “Clean Slate” Rule (Full Formatting): Before starting a new measurement campaign, it is critical not to reuse cards without fully formatting them first. Manually deleting files leaves hidden indexes and fragmentation. A clean format ensures that the microcontroller does not consume resources processing foreign directories (preventing packet loss due to write latency) and prevents the “cross-contamination” of data with previous projects.
Investing in “High Endurance” Cards: Standard memory cards suffer premature wear when recording data 24/7. For environmental dataloggers deployed outdoors and subject to significant thermal fluctuations, high-endurance cards (High/Pro Endurance) are essential. They are specifically designed to withstand thousands of hours of constant rewriting, shielding data against physical corruption.
Treating the SD card with scientific rigor from day one ensures we do not lose months of continuous sampling.