Exploring lag times between monthly atmospheric deposition and stream chemistry in Appalachian forests using cross-correlation

David R. DeWalle, Elizabeth Weeks Boyer, Anthony Buda

Research output: Contribution to journalArticle

Abstract

Forecasts of ecosystem changes due to variations in atmospheric emissions policies require a fundamental understanding of lag times between changes in chemical inputs and watershed response. Impacts of changes in atmospheric deposition in the United States have been documented using national and regional long-term environmental monitoring programs beginning several decades ago. Consequently, time series of weekly NADP atmospheric wet deposition and monthly EPA-Long Term Monitoring stream chemistry now exist for much of the Northeast which may provide insights into lag times. In this study of Appalachian forest basins, we estimated lag times for S, N and Cl by cross-correlating monthly data from four pairs of stream and deposition monitoring sites during the period from 1978 to 2012. A systems or impulse response function approach to cross-correlation was used to estimate lag times where the input deposition time series was pre-whitened using regression modeling and the stream response time series was filtered using the deposition regression model prior to cross-correlation. Cross-correlations for S were greatest at annual intervals over a relatively well-defined range of lags with the maximum correlations occurring at mean lags of 48 months. Chloride results were similar but more erratic with a mean lag of 57 months. Few high-correlation lags for N were indicated. Given the growing availability of atmospheric deposition and surface water chemistry monitoring data and our results for four Appalachian basins, further testing of cross-correlation as a method of estimating lag times on other basins appears justified.

Original languageEnglish (US)
Pages (from-to)206-214
Number of pages9
JournalAtmospheric Environment
Volume146
DOIs
StatePublished - Dec 1 2016

Fingerprint

atmospheric deposition
time series
basin
wet deposition
erratic
monitoring
environmental monitoring
water chemistry
chloride
watershed
surface water
ecosystem
modeling

All Science Journal Classification (ASJC) codes

  • Environmental Science(all)
  • Atmospheric Science

Cite this

@article{90452595dccc4d3aba0ee8e8ed86b907,
title = "Exploring lag times between monthly atmospheric deposition and stream chemistry in Appalachian forests using cross-correlation",
abstract = "Forecasts of ecosystem changes due to variations in atmospheric emissions policies require a fundamental understanding of lag times between changes in chemical inputs and watershed response. Impacts of changes in atmospheric deposition in the United States have been documented using national and regional long-term environmental monitoring programs beginning several decades ago. Consequently, time series of weekly NADP atmospheric wet deposition and monthly EPA-Long Term Monitoring stream chemistry now exist for much of the Northeast which may provide insights into lag times. In this study of Appalachian forest basins, we estimated lag times for S, N and Cl by cross-correlating monthly data from four pairs of stream and deposition monitoring sites during the period from 1978 to 2012. A systems or impulse response function approach to cross-correlation was used to estimate lag times where the input deposition time series was pre-whitened using regression modeling and the stream response time series was filtered using the deposition regression model prior to cross-correlation. Cross-correlations for S were greatest at annual intervals over a relatively well-defined range of lags with the maximum correlations occurring at mean lags of 48 months. Chloride results were similar but more erratic with a mean lag of 57 months. Few high-correlation lags for N were indicated. Given the growing availability of atmospheric deposition and surface water chemistry monitoring data and our results for four Appalachian basins, further testing of cross-correlation as a method of estimating lag times on other basins appears justified.",
author = "DeWalle, {David R.} and Boyer, {Elizabeth Weeks} and Anthony Buda",
year = "2016",
month = "12",
day = "1",
doi = "10.1016/j.atmosenv.2016.09.015",
language = "English (US)",
volume = "146",
pages = "206--214",
journal = "Atmospheric Environment",
issn = "1352-2310",
publisher = "Elsevier Limited",

}

Exploring lag times between monthly atmospheric deposition and stream chemistry in Appalachian forests using cross-correlation. / DeWalle, David R.; Boyer, Elizabeth Weeks; Buda, Anthony.

In: Atmospheric Environment, Vol. 146, 01.12.2016, p. 206-214.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Exploring lag times between monthly atmospheric deposition and stream chemistry in Appalachian forests using cross-correlation

AU - DeWalle, David R.

AU - Boyer, Elizabeth Weeks

AU - Buda, Anthony

PY - 2016/12/1

Y1 - 2016/12/1

N2 - Forecasts of ecosystem changes due to variations in atmospheric emissions policies require a fundamental understanding of lag times between changes in chemical inputs and watershed response. Impacts of changes in atmospheric deposition in the United States have been documented using national and regional long-term environmental monitoring programs beginning several decades ago. Consequently, time series of weekly NADP atmospheric wet deposition and monthly EPA-Long Term Monitoring stream chemistry now exist for much of the Northeast which may provide insights into lag times. In this study of Appalachian forest basins, we estimated lag times for S, N and Cl by cross-correlating monthly data from four pairs of stream and deposition monitoring sites during the period from 1978 to 2012. A systems or impulse response function approach to cross-correlation was used to estimate lag times where the input deposition time series was pre-whitened using regression modeling and the stream response time series was filtered using the deposition regression model prior to cross-correlation. Cross-correlations for S were greatest at annual intervals over a relatively well-defined range of lags with the maximum correlations occurring at mean lags of 48 months. Chloride results were similar but more erratic with a mean lag of 57 months. Few high-correlation lags for N were indicated. Given the growing availability of atmospheric deposition and surface water chemistry monitoring data and our results for four Appalachian basins, further testing of cross-correlation as a method of estimating lag times on other basins appears justified.

AB - Forecasts of ecosystem changes due to variations in atmospheric emissions policies require a fundamental understanding of lag times between changes in chemical inputs and watershed response. Impacts of changes in atmospheric deposition in the United States have been documented using national and regional long-term environmental monitoring programs beginning several decades ago. Consequently, time series of weekly NADP atmospheric wet deposition and monthly EPA-Long Term Monitoring stream chemistry now exist for much of the Northeast which may provide insights into lag times. In this study of Appalachian forest basins, we estimated lag times for S, N and Cl by cross-correlating monthly data from four pairs of stream and deposition monitoring sites during the period from 1978 to 2012. A systems or impulse response function approach to cross-correlation was used to estimate lag times where the input deposition time series was pre-whitened using regression modeling and the stream response time series was filtered using the deposition regression model prior to cross-correlation. Cross-correlations for S were greatest at annual intervals over a relatively well-defined range of lags with the maximum correlations occurring at mean lags of 48 months. Chloride results were similar but more erratic with a mean lag of 57 months. Few high-correlation lags for N were indicated. Given the growing availability of atmospheric deposition and surface water chemistry monitoring data and our results for four Appalachian basins, further testing of cross-correlation as a method of estimating lag times on other basins appears justified.

UR - http://www.scopus.com/inward/record.url?scp=84994378342&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84994378342&partnerID=8YFLogxK

U2 - 10.1016/j.atmosenv.2016.09.015

DO - 10.1016/j.atmosenv.2016.09.015

M3 - Article

AN - SCOPUS:84994378342

VL - 146

SP - 206

EP - 214

JO - Atmospheric Environment

JF - Atmospheric Environment

SN - 1352-2310

ER -