We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Predicting Daily River Chlorophyll Concentrations at a Continental Scale.
- Authors
Savoy, Philip; Harvey, Judson W.
- Abstract
Eutrophication is one of the largest threats to aquatic ecosystems and chlorophyll a measurements are relevant indicators of trophic state and algal abundance. Many studies have modeled chlorophyll a in rivers but model development and testing has largely occurred at individual sites which hampers creating generalized models capable of making broad‐scale predictions. To address this gap, we compiled a large data set of chlorophyll a concentrations matched to other water quality, meteorological, and reach characteristic data for a diverse set of 82 streams and rivers across the United States. We used this data set and extreme gradient boosting, a tree‐based machine learning algorithm, to predict daily chlorophyll a concentrations. Furthermore, we tested several practical considerations of broad‐scale models, such as making predictions at sites not included in model training or the utility of in situ water quality data versus universally available remotely estimated model inputs. Predictions were very strongly correlated to observations when compared against a randomly withheld subset of days; however, the model had lower accuracy when applied to completely novel sites withheld from model training. Turbidity and total nitrogen were the two most important variables for predicting chlorophyll a. Although in situ variables improved modeled estimates and were identified as more important during model interpretation, using only remote inputs still resulted in highly correlated predictions with small bias. Testing a model across many sites allowed for identification of common variables relevant to chlorophyll a and highlighted several challenges for applying data‐driven models to new sites or at larger spatial scales. Plain Language Summary: One of the largest threats to freshwater and marine ecosystems is excess primary productivity stimulated by increased nutrient availability. An overabundance of algae can cause degradation of ecosystems and risks for public health. Chlorophyll a routinely collected as a proxy measure for algal abundance and provides a common currency to compare across aquatic ecosystems. We compiled a data set of chlorophyll a, additional in situ water quality variables, reach characteristics, and broadscale meteorology to model daily chlorophyll concentrations at 82 rivers across the United States. Machine learning was used to predict daily chlorophyll concentrations and several challenges of broad‐scale modeling were investigated, such as extrapolating predictions to new sites or tradeoffs between data quality and quantity. While the model performed well against a withheld set of days for testing, it performed noticeably worse when entire sites were withheld for model testing. Turbidity and total nitrogen were the two most important predictors, and although in situ data resulted in better predictions, estimates solely made from remote data sets have potential for use in widespread predictions. Testing a model across many sites allowed us to identify variables that are commonly relevant to chlorophyll a and highlighted needs for future improvement of broad‐scale models. Key Points: Predicted daily chlorophyll a concentrations across a diverse set of streams and rivers using a decision tree machine learning algorithmModel estimates performed well at sites that contributed some data to model training but not at sites totally withheld from model trainingUsing in situ data for model inputs increased accuracy but ubiquitous remote data sets can provide reasonable estimates at many more places
- Subjects
UNITED States; MULTISCALE modeling; CHLOROPHYLL; MACHINE learning; ECOSYSTEMS; MONETARY unions; WATER quality; MARINE ecology; DECISION trees
- Publication
Water Resources Research, 2023, Vol 59, Issue 11, p1
- ISSN
0043-1397
- Publication type
Article
- DOI
10.1029/2022WR034215