Predicting stream channel erosion in the lacustrine core of the upper Nemadji River, Minnesota (USA) using stream geomorphology metrics

Joseph A. Magner, Kenneth N. Brooks

Research output: Contribution to journalArticlepeer-review

8 Scopus citations

Abstract

The USA Clean Water Act requires the development of a total maximum daily load (TMDL) when Minnesota's water quality standard for turbidity is exceeded; however, regions underlain with fine-grained lacustrine deposits yield large natural background loads of suspended inorganic sediment. A review of hydrogeologic pathways was conducted along with the statistical analysis of geomorphic metrics, collected at 15 sites with varying drainage areas in the upper Nemadji River basin, northeastern Minnesota. Regression analysis indicated a strong linkage between bankfull cross-sectional area and drainage area. Dimensionless geomorphic metric ratios were developed to predict channel evolution potential and associated channel erosion risk. Sites located in drainage areas less than 2 km2 had low erosion risk and showed a correlation between channel slope and relative roughness (D 84/mean bankfull channel depth, 88%). A principal components analysis explained over 98% of the variance between sites and indicated five important channel shape metrics to predict channel erosion: bankfull width, bankfull depth, maximum depth, cross-sectional area, and valley beltwidth. Mass wasting of cohesive stream channel sediment was influenced by groundwater discharge and produced turbid waters in the upper Nemadji River.

Original languageEnglish (US)
Pages (from-to)1423-1434
Number of pages12
JournalEnvironmental Geology
Volume54
Issue number7
DOIs
StatePublished - Jun 1 2008

Keywords

  • Channel stability
  • Cohesive-sediment
  • Erosion risk
  • Turbidity and water quality

Fingerprint Dive into the research topics of 'Predicting stream channel erosion in the lacustrine core of the upper Nemadji River, Minnesota (USA) using stream geomorphology metrics'. Together they form a unique fingerprint.

Cite this