Environ ResJClub
Data-driven approach to estimate ambient ammonia concentrations.
Environmental Research · · Journal Article · Open access
Govande, Heederik + more
Abstract ↗AI summary
The abstract is read at the publisher; the summary is JClub's.
Linear regression best predicted one-day-ahead ambient ammonia (NH 3 ) at most of six Dutch sites, reaching R 2 = 0.84 at Zegveld after correcting for autocorrelation lag.
- Why it matters: Reactive nitrogen deposition from NH 3 harms soils, waters, and biodiversity, and instrument failures can create monitoring gaps in intensive livestock regions. Reliable gap-filling models are needed to maintain continuous concentration estimates.
- What they did: They trained Linear Regression, Random Forest, and GRU models to forecast one day ahead NH 3 at De Zilk, Valthermond, Vredepeel, Wekerom, Wieringerwerf, and Zegveld, using lagged NH 3 plus wind speed and relative humidity. After adjusting for prediction lag from autocorrelation, they evaluated performance and used linear regression for gap filling.
- The result: Linear regression maintained reasonable accuracy with sparse observations every 2–7 days, and most sites showed significant correlations of almost or more than 0.5 even at 7 days gaps. Historical data from existing sites could also fill gaps at newly established locations with similar NH 3 distributions.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.
- Open access
Free. We’ll email you when a new paper cites this one or any other paper you follow on JClub — at most one email a day. Every email has a one-tap unsubscribe.