Scholarlink

Local meteoric water lines (LMWLs) provide essential information about the isotopic characteristics of local precipitation, serving as a benchmark for studying various hydrological processes and moisture sources in an area. Once isotopic compositions (δ18O and δ2H) of precipitation are available, LMWLs are usually developed using ordinary least square regression (OLR), which assumes that the independent variable (δ18O) is error-free, while the response variable (δ2H) may contain errors. However, this assumption is invalid since both δ2H and δ18O have measurement uncertainties. In this study, Type A and Type B measurement uncertainties in δ2H and δ18O were quantified for 58 rainwater samples collected during 2022–23 at an experimental plot in IIT Kanpur. LMWLs were derived using major axis (MA) and reduced major axis (RMA) regression methods that relax the assumption of error-free independent variable in regression analysis. A Monte Carlo based OLR method was developed that can use known (but varying) measurement errors in the regression analysis. The slope and intercept of LMWL from OLR (δ2H = 7.84 ± 0.15 δ18O + 8.27 ± 2.65, R2 = 0.98) were different from Monte Carlo OLR (δ2H = 7.96 ± 0.09 δ18O + 9.73 ± 0.80, R2 = 0.98). A comparison of OLR, MA and RMA showed that if measurement error information is not available, MA estimates LMWL parameters are closer to Monte Carlo OLR. Therefore, MA method is recommended for constructing LMWL.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Development of Local Meteoric Water Line for Kanpur, Uttar Pradesh

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • P. Singh,
  • D. Balan,
  • R. Ojha,
  • R. Srivastava,
  • S. Tripathi,
  • S. Guha,
  • G. Krishan,
  • MS. Rao,
  • P. Thakur

摘要

Local meteoric water lines (LMWLs) provide essential information about the isotopic characteristics of local precipitation, serving as a benchmark for studying various hydrological processes and moisture sources in an area. Once isotopic compositions (δ18O and δ2H) of precipitation are available, LMWLs are usually developed using ordinary least square regression (OLR), which assumes that the independent variable (δ18O) is error-free, while the response variable (δ2H) may contain errors. However, this assumption is invalid since both δ2H and δ18O have measurement uncertainties. In this study, Type A and Type B measurement uncertainties in δ2H and δ18O were quantified for 58 rainwater samples collected during 2022–23 at an experimental plot in IIT Kanpur. LMWLs were derived using major axis (MA) and reduced major axis (RMA) regression methods that relax the assumption of error-free independent variable in regression analysis. A Monte Carlo based OLR method was developed that can use known (but varying) measurement errors in the regression analysis. The slope and intercept of LMWL from OLR (δ2H = 7.84 ± 0.15 δ18O + 8.27 ± 2.65, R2 = 0.98) were different from Monte Carlo OLR (δ2H = 7.96 ± 0.09 δ18O + 9.73 ± 0.80, R2 = 0.98). A comparison of OLR, MA and RMA showed that if measurement error information is not available, MA estimates LMWL parameters are closer to Monte Carlo OLR. Therefore, MA method is recommended for constructing LMWL.