Date of Award

12-2012

Degree Type

Thesis

Degree Name

Master of Science (MS)

Department

Civil and Environmental Engineering

Advisor(s)

David Chandler

Keywords

Catskill, solute flux, ungaged, watershed hydrology

Subject Categories

Environmental Engineering

Abstract

Predictions of flow and subsequent solute fluxes from ungaged basins have important implications both for water resources management and ecosystem monitoring studies. The Catskill region of New York State is one such place that requires both water resources management and ecosystem monitoring due to its strategic location as the main water-supplying region for New York City. This study examines the differences in chemical mass flux estimates made in ungaged basins using three different chemistry aggregation methods for solute concentrations determined from monthly grab samples. The efficacy of area ratios for predicting flow at the upstream location of a nested pair of stream gages based on flow at the downstream reference gage is also explored. The benefit of data set partitioning and development of separate prediction models for different flow regimes of the reference gage is analyzed, and a threshold of area ratio for use of such a method is established, with implications for use in ungaged basins. This work is focused on the Catskill region, but is likely to be applicable to other temperate, montane systems.

Significant relationships were observed between upstream and downstream flow in all test watersheds. Furthermore, watershed area ratio was the most important basin parameter for estimating flow at the upstream location of a nested pair of stream gages. The area ratio alone explained 93% of the variance in the functional relation slopes that best fit the flow regressions. Data set partitioning was found to be beneficial only for nested pairs with area ratios greater than 0.1, and was determined by analysis of the root mean square error of the different flow prediction models. Five of the fifteen test watershed pairs had a lower root mean square error using the partitioned relationships and these pairs all had area ratios greater than 0.1.

The relative difference between the three different chemistry aggregation methods was found to be relatively small on an annual basis (average difference of 7%) and increase with shorter time steps up to daily flux estimates (average difference of 26%). This finding indicates that simple flow estimation methods based on area ratios are justifiable, and perhaps preferred, for estimation of annual chemical mass fluxes, and that for such estimates of flux, the exact solute chemistry aggregation method matters little on an annual basis.

Access

Open Access

Download

Share

COinS