.png){kind=icon}
.png){kind=icon}
Return
A Quantitative Basis for Intersource Differences in Tracer Concentrations for Accurate Sediment Source Fingerprinting
.png){kind=icon}
PRE
AI
DOI：10.1002/ldr.5490.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Sediment source fingerprinting provides support for targeted soil erosion control and land degradation mitigation. Tracer selection and source classification are the main considerations in this approach. However, the impact of tracer concentration divergences among different sediment sources on the accuracy of source apportionment remains uncertain and has not been quantitatively investigated. Here, artificial and virtual mixtures were created to explore the impact of low and high source differences on the accuracy of relative source apportionment. Additionally, global geochemical property data based on different source classification methods were collected to analyze the reliability of source classification. The results reveal that as the difference in tracer concentration between sources decreases (from 30% to 2%), the accuracy of source apportionment significantly decreased, with the mean absolute error (MAE) ranging from 5.9% to 60.1%. Furthermore, a relative deviation in tracer concentration > 10% between sources is suggested for reliable source classifications. Notably, approximately 40% of the existing papers exhibit potential issues due to low tracer concentration divergences between sources. The adoption of sediment source classification based on land use should be cautiously approached until tracer concentration divergences among major sources are examined. For catchments lacking long-term stable vegetation, source classification methods based on geological properties are recommended. Finally, we recommend the consideration of clustering methods or alterative tracers for specific study areas to improve the accuracy of sediment fingerprinting.
Keywords:
artificial mixtures
sediment fingerprinting
soil erosion
source classifications
unmixing modeling
virtual mixtures
Journal
IF:
3.7
Papers:
4.3K
Citations:
1.3W
View more organizations