AmericanCongressonSurveyingandMapping(ACSM),isthescienceandartofmakingallessentialmeasurementstodeterminetherelativepositionofpointsorphysicalandculturaldetailsabove,on,orbeneaththesurfaceoftheEarth,andtodepicttheminausableform,ortoestablishthepositionofpointsordetails.commonlypracticedbysurveyors.Abasictenetofsurveyingisthatnomeasurementisperfect,andthattherewillalwaysbeasmallamountoferror.Surveyerrorsareclassedintothreetypes:Grosserrorsorblunders:areerrorsmadebythesurveyorduringthesurvey,forexample,upsettingtheinstrument,misaimingatatarget,orwritingdownameasurementwrong.Alargegrosserrormayreducetheaccuracytoanunacceptablelevel.Thereforesurveyorsuseredundantmeasurementsandindependentcheckstodetecttheseerrorsearlyinthesurvey.Systematicerrors:areerrorsthatfollowaconsistentpattern.ExamplesincludeeffectsoftemperatureonachainorEDMmeasurement,orapoorlyadjustedspirit-levelresultinginamisalignedinstrumentortargetpole.Systematicerrorsthathaveknowneffectscanbecompensatedorcorrected.Randomerrors:Randomerrorsarethesmallunavoidablemeasurementscausedbyimperfectionsinmeasuringequipment,eyesightandconditions.Theycanbeminimisedbyredundancyofmeasurementandavoidingunstableconditions.Randomerrorstendtocanceleachotherout,butchecksmustbemadetoensuretheyarenotpropagatingfromonemeasurementtothenext.Surveyorsusetoolssuchastotalstations,robotictotalstations,GPSreceivers,prisms,3Dscanners,radios,handheldtablets,digitallevels,andsurveyingsoftware.Adjustmentofobservationsisamodelforthesolutionofanoverdeterminedsystemofequationsbasedontheprincipleofleastsquares.Itisusedextensivelyinthedisciplinesofsurveying,geodesy,photogrammetry(thefieldofgeomatics,collectively).Therearethreeformsofadjustmentofobservations:parametric,conditional,andcombined.Inparametricadjustment,onecanfindanobservationequationh(X)=YrelatingobservationsYexplicitlyintermsofparametersX(leadingtotheA-modelbelow).Inconditionaladjustment,thereexistsaconditionequationg(Y)=0involvingonlyobservationsY(leadingtotheB-modelbelow),withnoparametersXatall.Finally,inacombinedadjustment,bothparametersXandobservationsYareinvolvedimplicitlyinamixed-modelequationf(X,Y)=0.Clearly,parametricandconditionaladjustmentscorrespondtothemoregeneralcombinedcasewhenf(X,Y)=h(X)-Yandf(X,Y)=g(Y),respectively.Yetthespecialcaseswarrantsimplersolutions,asdetailedbelow.Oftenintheliterature,YmaybedenotedL.Themethodofleastsquaresisastandardapproachtotheapproximatesolutionofoverdeterminedsystems,i.e.,setsofequationsinwhichtherearemoreequationsthanunknowns."Leastsquares"meansthattheoverallsolutionminimizesthesumofthesquaresoftheerrorsmadeintheresultsofeverysingleequation.Themostimportantapplicationisindatafitting.Thebestfitintheleast-squaressenseminimizesthesumofsquared

residuals,aresidualbeingthedifferencebetweenanobservedvalueandthefittedvalueprovidedbyamodel.Whentheproblemhassubstantialuncertaintiesintheindependentvariable(the'x'variable),thensimpleregressionandleastsquaresmethodshaveproblems;insuchcases,themethodologyrequiredforfittingerrors-in-variablesmodelsmaybeconsideredinsteadofthatforleastsquares.Leastsquaresproblemsfallintotwocategories:linearorordinaryleastsquaresandnon-linearleastsquares,dependingonwhetherornottheresidualsarelinearinallunknowns.Thelinearleast-squaresproblemoccursinstatisticalregressionanalysis;ithasaclosed-formsolution.Aclosed-formsolution(orclosed-formexpression)isanyformulathatcanbeevaluatedinafinitenumberofstandardoperations.Thenon-linearproblemhasnoclosed-formsolutionandisusuallysolvedbyiterativerefinement;ateachiterationthesystemisapproximatedbyalinearone,andthusthecorecalculationissimilarinbothcases.Unit1

TheReliabilityofaSurveyandErrorsNew

WordsandExpressionsreliability[rilai[5biliti]n.

可靠性error[5er[]n.

誤差，錯誤issubjectto：受影響的accuracy[5Akjur[si]

n.

精度（有時特指最后結(jié)果的精度）accuracyrequired：要求的精度thescaleoftheplot：繪圖比例尺plottableerror：繪圖誤差draughtsman[5drAftsm[n]n.

制圖員，繪圖員represent[repri5zent]vt.

表現(xiàn)，描繪，象征plottabledistance：可繪距離specification[spesifi5keiF[n]n.

規(guī)格，規(guī)范，技術(shù)要求，技術(shù)條件；說明書，明細表tolerance[5tCl[r[ns]n.

限差，容許誤差；公差standard[5stAnd[d]n.

標準，規(guī)范angle[5ANg[l]n.

角，角度distance[5dist[ns]n.

距離，長度，間距salient[5seili[nt]a.

易見的，顯著的，突出的，跳躍的blunder[5blQnd[]n.

粗差，錯誤grosserror：粗差transpose[trans5p[uz]vt.

vi.

改變順序，互換位置well-planned：精心設計的observingprocedure：觀測程序systematicerror：系統(tǒng)誤差expansion[ik5spAnF[n]n.

擴充，膨脹frequency[5fri:kw[nsi]n.頻率，周期，發(fā)生次數(shù)ElectromagneticDistanceMeasuring(EDM)：電磁測距儀collimation[kCli5meiF[n]n.

瞄準consist

of：包括succession[s[k5seF[n]n.

連續(xù)，演替level(l)ing[5lev[liN]n.

水準測量；抄平，置平traverse[5trAv[:s]

n.導線，導線測量；v.

橫移，交叉traversing[5trAv[:siN]v.導線測量neglected[nig5lektid]a.被忽視的resultin：導致eliminate[i5limineit]vt.

排除，消除correction[k[5rekF[n]n.校正，修正，；修正量，改正數(shù)instrument[5instrum[nt]n.儀器，器械，器具，工具；手段，方法randomerror：隨機誤差，偶然誤差identical[ai5dentik[l]a.一致的，同樣的normally

distributed：正態(tài)分布的normaldistribution：正態(tài)分布dealwith：安排，處理，涉及accumulationoferror：誤差累積workingfromthewholetothepart：從整體到局部redundant

observations：多余觀測triangle[5traiAngl]n.三角形precision[pri5siV[n]n.精確，精密度，精度independent[indi5pend[nt]a.獨立的inspect[in5spekt]vt.

檢查，視察NotestotheText1．Agooddraughtsmancanplotalengthtowithin0.25mmandso,ifaplanofanareaisrequiredatascaleof1/1000,i.e.,1mmontheplanrepresents1montheground,thesmallestplottabledistanceis0.25m.Thus,forasurveyat1/1000scale,allthemeasurementsmustbetakensuchthattherelativepositionsofanypointwithrespecttoanyothermustbedeterminedto0.25morbetter.譯文：一個好的繪圖員可以在大約0.25mm的范圍內(nèi)畫圖，假如一個地區(qū)的圖紙比例尺要求是1/1000，即1mm圖上距離代表1m實地距離，那么最小的測距應該是0.25m。因此，對于1/1000比例尺的測量，所有測量必須保證任意一點與其他任意點的相對位置不得大于0.25m。2．Thereareseveraltypesoferrorthatoccurandaknowledgeoftheirimportanceandcharacteristicsisessentialintheunderstandingofthelimitationsofthetechniquesofmeasurement.Thesalientfeatureswillbestatednowbutfurtherinformationisgiveninlaterchapters.譯文：在測量中有幾種類型的誤差，掌握它們的重要性和特性對于理解測量技術(shù)的限差是十分必要的。本章主要講解這些誤差的主要特征，其后的章節(jié)會對它們做更深入的闡述。3．Theseerrorsareofvitalimportanceinactivitieswhichconsistofaddingtogetherasuccessionofindividualobservations(seesectionsonlevelingandtraversing).Ifalltheindividualmeasurements