SECONDARY VALIDATION RAINFALL DATA PRIMARY VALIDATION ALREADY DONE

SECONDARY VALIDATION - RAINFALL DATA • PRIMARY VALIDATION ALREADY DONE * ON INDIVIDUAL STATION BASIS • SECONDARY VALIDATION * IDENTIFY SUSPECT VALUES BY HAVING COMPARISON WITH NEIGHBOURING STATIONS • DIVISIONAL DPCs * TO CONTINUE AT SDPCs • RAINFALL NOT HIGHLY CORRELATED IN SPACE & TIME !! * NOT A ONE-TO-ONE - COMPARISON DIFFICULT * BETTER CORRELATION - LARGER TIME INTERVALS * MISTAKES & SUSTAINED SYSTEMATIC ERRORS OHS - 1

SPATIAL CORRELATION • MAJOR INFLUENCING FACTORS – – OHS - 2 LEVEL OF AGGREGATION IN TIME DISTANCE TYPE OF PRECIPITATION PHYSIOGRAPHIC FACTORS

KHEDA CATCHMENT - SABARMATI OHS - 3

OHS - 4

OHS - 5

OHS - 6

OHS - 7

REGION A OHS - 8

OHS - 9

EXTENDED REGION OHS - 10

USE OF SPATIAL CORRELATION • PREVAILING SPATIAL CORRELATION – AS BASIS FOR TESTING SPATIAL CONSISTENCY – SPATIAL INTERPOLATION • THERE IS LIKELIHOOD OF – REJECTING GOOD BUT ANAMOLOUS DATA – ACCEPTING BAD DATA • RELIANCE ON STATION REPUTATION – TYPE OF STATION – PAST RECORD ON QUALITY – OBSERVER SINCERITY & PAST RECORDS OHS - 11

SECONDARY DATA VALIDATION • SCREENING OF DATA • MULTIPLE TIME SERIES GRAPHS • DAILY RAINFALL TABULATION FOR NEIGHBOURING STATIONS • APPLYING DATA LIMITS AT LONGER DURATIONS • SPATIAL HOMOGENEITY TEST • VERY COMMON ERRORS – ACCUMULATION OF RAINFALL OR MISSED ENTRY – MISSED RAINFALL ON LOW RAINFALL DAYS • DOUBLE MASS ANALYSIS OHS - 12 – LONG TERM SYSTEMATIC SHIFTS

DATA VALIDATION • SCREENING OF DATA – SCREENING AGAINST DATA LIMITS – TABULAR OUTPUT – BASIC STATISTICS • OBJECTIVES – REVIEW PRIMARY VALIDATION – HARD COPY * MEDIUM FOR SCRIBBLING REMARKS/ DATA VALIDATION NOTES OHS - 13

OHS - 14

OHS - 15

OHS - 16

SCRUTINY BY TABULATION OHS - 17

OHS - 18

SPATIAL HOMOGENEITY TESTING • SPATIAL CORRELATION – CONSISTENCY BETWEEN NEIGHBOURING STATIONS – COMPARISON OF * OBSERVED VALUE * VALUE ESTIMATED ON THE BASIS OF NEARBY STATIONS OHS - 19

TEST & BASE (NEIGHBOUR) STATIONS + II + + + Test station + + OHS - 20 + + Neighbouring stations + Selected neighbouring + stations II etc. + + + III I + + IV Quadrant

• SPATIAL ESTIMATE - WEIGHTED AVERAGE – WEIGHTS : INVERSE OF DISTANCE POWER – TEST : ABSOLUTE & RELATIVE OHS - 21

OHS - 22

OHS - 23

COMMON ERRORS - SHIFT IN ENTRIES OHS - 24

OHS - 25

COMMON ERRORS - ACCUMULATIONS • RAINFALL OBSERVATIONS – TO BE MADE DAILY WITHOUT FAIL – ABSENCE DUE TO PRACTICAL REASONS * SICKNESS * HOLIDAY(S)/WEEKENDS – OBSERVERS TENDENCY * DISTRIBUTE RAINFALL BASED ON OWN GUESS * RECORD ACCUMULATED VALUE ON THE LAST DAY WITHOUT MENTIONING IT * RECORD ACCUMULATED VALUE ON THE LAST DAY WITH PROPER REMARKS (NOT PRACTICED NORMALLY) • DETECTION OF SUCH ERRORS – USING RECORDS OF NEARBY STATIONS OHS - 26

OHS - 27

OHS - 28

COMMON ERRORS - MISSED ENTRIES OHS - 29

OHS - 30

DOUBLE MASS ANALYSIS • DETECTION OF SYSTEMATIC SHIFTS IN TIME – PERSISTENCE FOR LONG TIME PERIODS – ABRUPT OR GRADUAL CHANGE • TEMPORAL INCONSISTENCY: FACTORS – – – SHIFTING OF STATION TO ANOTHER LOCATION CHANGE IN EXPOSURE CONDITION CHANGE IN EQUIPMENT OR PRACTICES UNATTENDED FAULT IN EQUIPMENT CHANGE IN REGIME: MAN-MADE TOO • IN-HOMOGENEITY IN DATA SET – NOT ACCEPTABLE IN STATISTICAL ANALYSIS – RELATING SERIES TO DESIRED SITE/REGIME OHS - 31

DETECTION OF INHOMOGENEITY • COMPARING BEHAVIOUR OF TEST STATION WITH RESPECT TO NEARBY STATIONS – BASE STATION(S) ASSUMED TO HOMOGENEOUS – ANY SIGNIFICANT & PERSISTENT DIFFERENCES IN MUTUAL RELATIONSHIP INDICATES SHIFT – RANDOM VARIATION IN RELATIONSHIP TEND TO COMPENSATE IF AGGREGATED IN TIME – SYSTEMATIC VARIATION TENDS TO SHOW UP OHS - 32

DOUBLE MASS CURVE • SIMPLE TEST PROCEDURE – ACCUMULATE IN TIME * OBSEREVED TEST SERIES * ESTIMATED BASE SERIES – COMPARE THE TWO ACCUMULATED SERIES – SIGNIFICANT & PERSISTENT DEVIATION SHOWS IN-HOMOGENEITY IN THE TEST SERIES OHS - 33

OHS - 34

OHS - 35