peer review report 1 on scale dependence of controls on groundwater vulnerability in the water energy food nexus california coastal basin aquifer system

Comments to Author: clarification... 58 Peer Review Report / Journal of Hydrology: Regional Studies 9 2017 57–58Groundwater-qualitydatasectiononp.8:Aspertheabove,wellsamplingdatesshouldbe

Journal of Hydrology: Regional Studies 9 (2017) 57–58

Journal of Hydrology: Regional

Studies

j o u r n a l h o m e p a g e :w w w e l s e v i e r c o m / l o c a t e / e j r h

Peer review report 1 on “Scale dependence of controls on groundwater vulnerability in the water-energy-food nexus, California Coastal Basin aquifer system”

1 Original Submission

2 Comments to Author:

clarification

DOI of the original article: http://dx.doi.org/10.1016/j.ejrh.2016.01.002

http://dx.doi.org/10.1016/j.ejrh.2016.11.043

58 Peer Review Report / Journal of Hydrology: Regional Studies 9 (2017) 57–58

Groundwater-qualitydatasectiononp.8:Aspertheabove,wellsamplingdatesshouldbegivenabouthere.Isawno mentionoftheminthepaperorthesupplementarymaterial.Withoutthemitisitisdifficulttoevaluatetheappropriateness

oftheyearsoflanduseandfertilizerdatausedinthemodeling

L180–183?Wouldexpectamorelocallyinformedestimateofbackgroundnitrateintheregion,vs.usingnationalnumbers WhatisbackgroundforthispartofCA?SeemsthatsuchdatawouldbeavailablegivenCA’scomprehensiveGroundwater AmbientMonitoringandAssessmentprogram

Indicateaboutherethatthisisthelogisticregressionmodelingthreshold

L202–205:Mentiondatesoflanduseheresothereaderdoesn’thavetogotothesupplementalmaterial.Whynotuse

CADWRlandusedatainsteadofanationallandcoverdataset?Howdoesthelanduseyearcomparewithwellsampling dates−isthereasufficienttemporaloffsettoaccountforsystemlag?

L225:BecauseLRis“widelyused,¨suggestincludingmorereferenceshere

L242:Mentionherethat“univariaterelations”referstounivariatelogisticregressionmodels

L254–256:NeedsomedescriptionoftheLRfitmethodused.Forwardstepwise,backward,bestsubsets,or ?

L276:Suggestchanging“NO3”to“nitrogen ¨Notallfertilizercontainsnitrate

L278:Change“slow”to“slows”

L283–285:Itwouldbebettertotestformulticollinearitydirectlyinamultivariatemodelratherthanassumeitisnot occurringbasedonunivariatecorrelations.SeeMenard(2001,AppliedLogisticRegression,SagePublications)foruseofthe tolerancestatisticinalogisticregressioncontext

L309–325:AplausibleexplanationisprovidedfornegativerelationsbetweenNsourcevariablesandgroundwaternitrate

inthecentralsubregion.Traveltimefromthelandsurfacetothewellisanimportantfactorinthisregard,butthereisnot

alottogoonaspertheabovecommentsonwellsamplingdate

Inmodelb,DOisbasicallyjustaproxyfornitrate.AstheauthorsstateonL305-306,therelationbetweenDOandnitrate

iswelldocumented.Withoutasourceterm,thismodelisnotavulnerabilitymodel

Didyouconsiderinteractionsamongexplanatoryvariablestotryandteaseoutmoreprocesses?

L340–342:Seeabovecommentontestingformulticollinearityinalogisticregressionmodel

L353–354:Couldmentiontheimportanceofgroundwaterageabouthereaswell.Althoughallofthesesamplescontain somemoderngroundwateraccordingtothetritiumdata,theagemixturescouldvaryalotdependingonwelldepth,screen length,andpumpingrate

L356–357:Notquitesurewhatismeantby“scaleinvariant”forDO.Becauseitispresentinfourofthefivemodels? L363–365:SamplingforFeandMnalongwithDOwouldprovideevenmoreunderstandingofambientredoxconditions,

aspertheMcMahonandChapelle(2008)paper

L373:ExplainwhatAWCrepresentsabouthere

L376–379:Farmfertilizerisintheregionalmodelandnorthmodel,butnotthecentralandsouthmodels.Whyisit considered“scaleinvariant? ¨Thisvariableissaidtobe“leastsignificant”intheregionalmodel,butsoilAWChasahigherp valueaccordingtoTable3

L391–396:Notclearwhat“methods”youarereferringto.YousimplyassumedthesameDOvalueseverywheretomake themaps,correct?HowdotheselectedDOvaluesrelatetoredoxstatusasdescribedbyMcMahonandChapelle?

L411–413:Predictednitrateprobabilitiesseemratherlowandthereisnotalotofdifferencebetweenthetwomaps Maybeyoucouldmakeamapofthedifferences

“Predictiveability”sectionstartingonL416:Thisisnotmodelvalidation,thisisrecalibrationofamodeltodifferent data.Bydefinition,forittobepredictivetheobservationscan’thavebeenusedinmodelcalibration.Theauthorsimplythat simulatedvaluescan’tbecomparedwithobservedbecausetheformerareinprobabilityunits,butinfactitis straightfor-wardtocomparesimulatedandobservedoccurrencesandnon-occurrences(1sand0s);seeforexample“ClassifyingNew Observations”in“LogisticRegressionExamplesUsingtheSASSystem”(1997,SASInstituteInc.).Or,theycouldcompute deviancevaluesforthevalidationwellsbasedonthepredictedprobabilitiesandobserved0sand1s.Thisentiresection shouldberevisedordropped

L450:Whatisthetimeperiod?

L462–463:Farmfertilizerisdescribedas“scaleinvariant”butismissingfromthecentralandsouthmodels

Anonymous