What’s one piece of advice you wish you’d received before your first watershed-monitoring project?

How did pre-season equipment calibration (or lack thereof) affect your initial data quality, and what best practice would you now insist upon?