Graduate Thesis Or Dissertation
 

Forest Succession and Climate Change Effects on Long-Term Runoff Coefficients at Varying Timescales

Public Deposited

Downloadable Content

Download PDF
https://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/zg64ts859

Descriptions

Attribute NameValues
Creator
Abstract
  • Forest harvesting and climate change induced shifts in precipitation characteristics (i.e., intensity, type) may affect how water is partitioned on the landscape, with more water potentially being lost to evaporation or groundwater, and less water being partitioned into the stream. Long-term, paired watershed data allows us to assess these possible changes on the storm event scale and see the compounding effect of storms on the seasonal and annual timescales. Analysis of 50 years of storm-based, seasonal runoff coefficients (quickflow/storm precipitation) for five different paired-basin experiments in Western Oregon illustrates the seasonal variation in hydrologic response to harvest and the long-term time scales needed for streams to reach recovery. In the fall (September–November), the runoff coefficients in the treated watersheds were on average higher than those in the reference watersheds, with mean percent changes in runoff coefficients varying from 94 to 8%, up to 21–40 years after harvest. After this period of time, runoff coefficients in the treated watersheds declined relative to the reference watersheds, creating a streamflow deficit for the remainder of the record, with mean percent changes in runoff coefficients varying from -1 to -15%. In the winter season (December–February), in all cases except one, the runoff coefficients in the treated watersheds were greater than those in the reference watersheds for the entire record. The positive mean percent change in runoff coefficients varied from 38 to 5%. The magnitude of the positive mean percent change declined through time. The spring months (March–May) were similar to the winter months in that in all cases except one, the runoff coefficients in the treated watersheds were greater than those in the references (positive mean percent changes varying from 57 to 10%), and this difference declined through time. The spring months experienced a greater difference between the treated and reference watersheds than in the winter months. Using the shifts in storm-based, seasonal runoff coefficients as our base, we proposed different hydrological mechanisms that may be responsible for altering the volume of quickflow in the streams as forests regenerate. We outlined our discussion in terms of different temporary water storage pools (transpiration, canopy storage and evaporation, near surface storage, and snowpack storage) that may be absorbing incoming water and either slowing the transit time of the storm precipitation or causing water to be lost to evaporation. We suggested that these temporary water storage pools are changing in magnitude during forest succession, thereby changing the volume of water left to be partitioned into quickflow. We infer that transpiration will affect quickflow in the fall, while the effects of canopy storage and evaporation, near surface storage, and snowpack storage will be greatest in the fall and winter. We interpret these seasonal changes in terms of forest succession, but also note that these changes may compound up to the annual scale. Shifts in precipitation characteristics (i.e., intensity, type) caused by climate change have been documented throughout the Pacific Northwest and likely have an impact on how much water is partitioned into streamflow on the annual time scale. Analysis of over 50 years of daily precipitation and discharge data across ten watersheds indicate that annual runoff coefficients have declined in 70% of the watersheds, seemingly independent of any one physical property. To address what may be causing this widespread change, the following precipitation characteristics were calculated for five different storm magnitude designations (>= 5 and < 10 mm, >=10 and < 20 mm, >= 20 and < 30 mm, >= 30 and < 50 mm, and 50 mm or more) and averaged across each water year: maximum 15-minute rainfall intensity (mm/hr), mean rainfall intensity (mm/hr), mean number of days of no rainfall between storm events, mean storm duration (hrs), the mean total precipitation that fell seven days prior to the onset of a storm event, and mean total precipitation that fell fourteen days prior to the onset of a storm event. Additional data concerning snowpack accumulation and snowmelt were also calculated for each water year, including the calendar day of the year of maximum snow water equivalent, the maximum annual snow water equivalent (mm), and the maximum daily temperature between February and March. These data were then used in separate multiple linear regression models for each storm designation. Results indicated that the runoff coefficient in watersheds with younger forests is best modeled when using precipitation variables calculated for smaller storm designations, while the models that perform best for estimated runoff in watersheds with old growth forest are those that included larger storm designations variables. Antecedent precipitation variables that could be related to canopy interception processes were significant in small storm models, potentially indicating that these processes are most impactful for determining annual runoff coefficient values in the young forests. The maximum snow water equivalent positively was significantly correlated with annual runoff coefficients in 43 out of the 50 of the best-fit models. The strong correlations with snow water equivalent may indicate the importance of regional climatic trends in determining local annual runoff coefficients. We propose a two-component groundwater system that includes one pool of water that is mobile and one that is tightly bound and immobile. Our analyses indicate that annual runoff coefficients are not correlated with annual precipitation sums of previous years, suggesting that the mobile water is flushed out by the end of the dry season independent of climate. The volume of the immobile water may be determined by groundwater recharged of snowmelt at higher elevations. Declines in immobile water may lead to decreasing annual runoff coefficients like the trends observed here.
License
Resource Type
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Rights Statement
Publisher
Peer Reviewed
Language
Embargo reason
  • Ongoing Research
Embargo date range
  • 2020-05-21 to 2021-06-22

Relationships

Parents:

This work has no parents.

In Collection:

Items

Thumbnail Title Date Uploaded Visibility Actions
file details CrampeEmilyA2020.pdf 2020-05-21 Public Download