Contents lists available at ScienceDirect Journal of Geochemical Exploration journal homepage: www.elsevier.com/locate/gexplo Characterization of surface water isotope spatial patterns of Scotland Christian Birkel a,b, ⁎ , Rachel Helliwell c , Barry Thornton c , Sheila Gibbs c , Pat Cooper c , Chris Soulsby b , Doerthe Tetzlaff b , Luigi Spezia d , Germain Esquivel-Hernández e , Ricardo Sánchez-Murillo e , Andrew J. Midwood c,f a Department of Geography, University of Costa Rica, San José, P.O. Box 11501-2060, San José, Costa Rica b Northern Rivers Institute, University of Aberdeen, Aberdeen AB24 3UF, Scotland, United Kingdom of Great Britain and Northern Ireland c The James Hutton Institute, Aberdeen AB15 8QH, Scotland, United Kingdom of Great Britain and Northern Ireland d Biomathematics and Statistics Scotland, Aberdeen AB15 8QH, Scotland, United Kingdom of Great Britain and Northern Ireland e Stable Isotope Research Group, National University of Costa Rica, Heredia, Costa Rica f Department of Biology, University of British Columbia – Okanagan, Kelowna V1V 1V7, B.C., Canada ARTICLE INFO Keywords: Scotland Stable isotopes Isoscapes d-Excess Geographically weighted regression models ABSTRACT The extended National Waters Inventory of Scotland (NWIS) monitoring network in combination with an ex- tensive, supplementary low flow sampling campaign was used to create isoscapes of surface water for man- agement purposes at high spatial resolution (100 m grid) across Scotland. The δ 2 H isoscape shows a strong isotopic separation along a north-south and east-west topographic (mountainous to the north and west and lowlands to the east) and climatic (wetter west, drier east) gradients. Isotopes were enriched in the western domain and depleted in the east and central Highland domains. The surface water d-excess isoscape show more complex spatial variability mainly related to contrasting moisture sources (sub-tropical North Atlantic Ocean, the North Sea, Polar Continental, and the Arctic) as well as secondary evaporation processes. The two-year NWIS isotope record exhibited a significant seasonal evaporative effect on surface water isotopes that progresses from winter through to a maximum in autumn as indicated by Local Evaporation Lines (LELs). The surface water isoscapes can be efficiently reproduced with geographically weighted regression (GWR) models using gridded annual precipitation, remotely sensed actual evapotranspiration, land cover, soil wetness, catchment area, and mean elevation. The GWR models showed potential to assess isotopic changes under future climate and land use change. 1. Introduction The concept of “isotopic landscapes” or “isoscapes” (West et al., 2010) has become a valuable tool to understand hydro-climatic pro- cesses and their effect on water resources across different spatial scales (Bowen and Revenaugh, 2003; Terzer et al., 2013; Jasechko et al., 2013; Evaristo et al., 2015; Good et al., 2015). Historically, the use of stable water isotopes in the environmental sciences was dependent on analytical capability. Fortunately, the development of inexpensive in- strumentation based on laser spectroscopy has greatly enhanced our ability to characterise the temporal and spatial variability of isoscapes at high resolution (Gupta et al., 2009; Good et al., 2014; Munksgaard et al., 2014). Recently, isoscapes have emerged as a low-cost and ef- fective tracer visualization technique to understand precipitation dy- namics (Lachniet and Paterson, 2009; Sánchez-Murillo et al., 2016a, 2016b), groundwater recharge mechanisms (Heilweil et al., 2009; Jasechko and Taylor, 2015; Sánchez-Murillo et al., 2016a, 2016b), paleoclimate (Vimeux et al., 2005; Lachniet, 2009; Risi et al., 2010), and ultimately, enhance water resources management (Bowen and Good, 2015). However, to use isoscapes as a physically-based man- agement tool - as advocated by Bowen and Good (2015) - the spatial resolution needs to cover regional as well as local scales. The latter requires sampling efforts able to generate a complete spatial re- presentation of complex terrain. A sufficiently high density of ob- servations is particularly important in mountainous landscapes due to the complex interaction of hydro-meteorological processes and small scale landscape heterogeneity (Yamanaka et al., 2015). In addition to characterising spatial variation with the highest possible density monitoring network, the geostatistical method used for spatial interpolation is also critical. There are several algorithms available and all have advantages or disadvantages, but there is little guidance as to which algorithm is the most suitable for isoscape https://doi.org/10.1016/j.gexplo.2018.07.011 Received 9 July 2017; Received in revised form 29 April 2018; Accepted 17 July 2018 ⁎ Corresponding author at: Department of Geography, University of Costa Rica, San José, P.O. Box 11501-2060, San José, Costa Rica. E-mail address: christian.birkel@ucr.ac.cr (C. Birkel). Journal of Geochemical Exploration 194 (2018) 71–80 Available online 24 July 2018 0375-6742/ © 2018 Elsevier B.V. All rights reserved. T