DATA MANAGEMENT IN GIS n SPATIAL GEO DATA

DATA MANAGEMENT IN GIS n SPATIAL (GEO) DATA MODELS n NON SPATIAL (RELATIONAL) DATA MODELS n GEO-RELATIONAL MODEL By Prof. V L. Swaminathan

DATA COMPONENTS IN GIS Measured x-y Point Line Poly NON-LOCATIONAL (Description) Topological Relation Var Grid Soil 1. . . . . Net. Class Val Name Sand 1. 1 1. 2 Fine Coarse T 2 LOCATIONAL DATA (Spatial) T 1 GEOGRAPHIC DATA al n r po sio m n Te ime D

SPATIAL DATA MODELING REAL WORLD ELLIPSOID DIGITAL MODEL GIS DATABASE MODEL SPHEROID ANALOG MODEL MAP RASTER VECTOR

VECTOR SPAGHETTI STRUCTURE Y AXIS 15 31 11 6 X AXIS (0, 0) Feature No. Point 6 Line 11 Polygon 31 Polygon 15 Location x, y (Single Point) x 1, y 1; x 2, y 2; . . . xn, yn

VECTOR SPAGHETTI : Critique-1 Y AXIS 11 (0, 0) 15 31 6 X AXIS Feature No. Point 6 Line 11 Polygon 31 Polygon 15 Location x, y (Single Point) x 1, y 1; x 2, y 2; . . . xn, yn CAPTURES LOCATIONS ACCURATELY BUT: SHARED BOUNDARIES OF POLYGON 31 & 15 CAPTURED MORE THAN ONCE. IN REALITY SUCH REPETITIONS WILL BE MANY. PROBLEMS OF INTEGRITY, AND REDUNDANCY IN DATABASE

VECTOR SPAGHETTI : Critique-2 Y AXIS 11 (0, 0) 15 31 6 X AXIS Feature No. Point 6 Line 11 Polygon 31 Polygon 15 Location x, y (Single Point) x 1, y 1; x 2, y 2; . . . xn, yn POLYGON 31 IS INCLUDED IN 15. LINE 11 AND POINT 6 ARE ALSO WITHIN POLYGON 15. IN REALITY SUCH RELATIONSHIPS ARE MANY. SPATIAL RELATIONSHIPS HAVE TO BE COMPUTED GEOMETRICALL AND INFERRED IMPLICATIONS ON COMPUTATIONAL EFFICIENCY

VECTOR TOPOLOGY STRUCTURE 1 1 1 2 4 3 5 2 3 2 6 7 3 8 4 11 9 10 NODE-2 1 3 2 2 2 5 6 4 6 LINK POLYGON 5 6 4 LINK# R-POLY L-POLY NODE-1 1 1 0 3 2 2 0 4 3 2 1 3 4 1 0 1 5 3 2 4 6 3 0 2 7 3 3 5 8 4 3 6 9 5 4 7. . . . NODE 5 7 NODE# 1 2 3 4 5 6 7 X_COORD 23 17 29 26 8 22 24 Y_CORD 8 17 15 21 26 30 38

VECTOR TOPOLOGY STRUCTURE : CRITIQUE -1 1 2 4 1 1 3 5 2 3 2 6 4 7 3 8 NODE 5 POLYGON 5 6 4 LINK 11 9 10 7 n n CAPTURES CO-ORDINATES ONLY ONCE PROBLEMS OF REDUNDANCY AND INTEGRITY MINIMIZED LINK# R-POLY L-POLY NODE-1 NODE-2 1 1 0 3 1 2 2 0 4 3 3 2 1 3 2 4 1 0 1 2 5 3 2 4 2 6 3 0 2 5 7 3 3 5 6 8 4 3 6 4 9 5 4 7 6. . . . NODE# 1 2 3 4 5 6 7 X_COORD 23 17 29 26 8 22 24 Y_CORD 8 17 15 21 26 30 38

VECTOR TOPOLOGY STRUCTURE : CRITIQUE -2 1 2 4 1 1 3 5 2 3 2 6 4 7 3 8 NODE 5 POLYGON 5 6 4 LINK 11 9 10 7 LINK# R-POLY L-POLY NODE-1 NODE-2 1 1 0 3 1 2 2 0 4 3 3 2 1 3 2 4 1 0 1 2 5 3 2 4 2 6 3 0 2 5 7 3 3 5 6 8 4 3 6 4 9 5 4 7 6. . . . CAPTURES SPATIAL RELATIONSHIPS EXPLICITLY n n ASSOCIATION AMONGST THE SPATIAL FEATURES CONNECTIVITY

VECTOR TOPOLOGY STRUCTURE : CRITIQUE -3 1 2 4 1 1 3 5 2 3 2 6 4 7 3 8 NODE 5 POLYGON 5 6 4 LINK 11 9 10 7 LINK# R-POLY L-POLY NODE-1 NODE-2 1 1 0 3 1 2 2 0 4 3 3 2 1 3 2 4 1 0 1 2 5 3 2 4 2 6 3 0 2 5 7 3 3 5 6 8 4 3 6 4 9 5 4 7 6. . . . ASSOCIATION n n n FOR FINDING POLYGONS SURROUNDING 3, SEARCH THROUGH THE LIST LOOK FOR LINKS BOUNDING THE POLYGON-3 LOOK FOR ASSOCIATED R-POLY & L-POLY

VECTOR TOPOLOGY STRUCTURE : CRITIQUE -4 1 2 4 1 1 3 5 2 3 2 6 4 7 3 8 NODE 5 POLYGON 5 6 4 LINK 11 9 10 7 LINK# R-POLY L-POLY NODE-1 NODE-2 1 1 0 3 1 2 2 0 4 3 3 2 1 3 2 4 1 0 1 2 5 3 2 4 2 6 3 0 2 5 7 3 3 5 6 8 4 3 6 4 9 5 4 7 6. . . . CONNECTIVITY : NODE-7 & NODE-2 n n LOOK FOR LINKS EMNATING FROM NODE-7 i. e 9, 11, 10 FOCUS ON ONE LINK(i. e. 9) AND LOOK FOR ASSOCIATED NODES (i. e 6) LOOK FOR LINKS EMNATING FROM THIS NODE (i. e. 8, 7) CONTINUE THE PROCESS TILL NODE 2 IS REACHED FOR ALL THE PASSES

VECTOR TOPOLOGY STRUCTURE : CRITIQUE -5 1 2 4 1 1 3 5 2 3 2 6 4 7 3 8 NODE 5 POLYGON 5 6 4 LINK 11 9 10 7 n BY EXPLICITLY CAPTURING SPATIAL RELATIONSHIPS (i. e. TOPOLOGY u GEOMETRIC COMPUTATION PROBLEM IS TURNED INTO SEARCH PROBLEM LINK# R-POLY L-POLY NODE-1 NODE-2 1 1 0 3 1 2 2 0 4 3 3 2 1 3 2 4 1 0 1 2 5 3 2 4 2 6 3 0 2 5 7 3 3 5 6 8 4 3 6 4 9 5 4 7 6. . . . NODE# 1 2 3 4 5 6 7 X_COORD 23 17 29 26 8 22 24 Y_CORD 8 17 15 21 26 30 38

RASTER GRID STRUCTURE 2 1 3 4 File Structure: Row 1 1. 2 2 Column 1 2 Value 0 0. 1 1 0 1 1 1 0 0 Row 3 3. 4. . 0 0 0 1 1 1 1 0 1 1 0 0 1 1 1 0 0 0 1 1 1 0 0 0 Column 1 2. 1. . Value 1 1. 1. .

RASTER GRID : Critique-1 2 1 3 4 0 1 1 1 0 0 0 1 1 1 1 0 1 1 0 0 1 1 1 0 0 0 1 1 1 0 0 0 LINE AND POINT NOT REPRESENTED: SEPEARATE LAYERS REQUIRED FOR OVERLAPPING FEATURES POLYGONS, POINTS AND LINES

RASTER GRID : Critique-2 2 1 3 4 0’S AND 1’S REPEATED MANY TIMES UNNECESSARILY REDUNDANCY OF DATA STORAGE IMPLICATIONS ON STORAGE VOLUME 0 1 1 1 0 0 0 1 1 1 1 0 1 1 0 0 1 1 1 0 0 0 1 1 1 0 0 0 SOLUTION RUN LENGTH ENCODING ROW COLUMN VAL 1 1 -7 0 2 1 -6 1 2 7 0. . .

RASTER GRID : Critique-2 2 1 3 4 0 1 1 1 0 0 0 1 1 1 1 0 1 1 0 0 1 1 1 0 0 0 1 1 1 0 0 0 POLYGON 2 NOT REPRESENTED: GRID SIZE HAS TO BE SMALLER IN ORDER TO REPRESENT FEATURES ACCURATELY - IMPLICATIONS ON STORAGE VOLUME OR HAVE VARIABLE SIZE OF GRID OVERLAY : QUAD-TREE

QUAD-TREE STRUCTURE RASTER n n QUADTREE AREA RECURSIVELY DECOMPOSED INTO REGULAR SHAPED GRIDS OF SMALLER & SMALLER SIZE TILL SMALLEST GRID REPRESENTS HOMOGENEOUS REGION REQUIRES LESS STORAGE THAN RASTER AS NUMBER OF GRIDS WILL BE LESS GRIDS COULD BE TRIANGLES, SQUARES OR HEXAGON SQUARES COMMONLY USED BECAUSE ORIENTATION PROBLEM COMES IN FOR OTHER SHAPES

QUAD-TREE STRUCTURE : CRITIQUE-1 AB C D A n RESULTS IN TREE STRUCTURE (PARENT CHILD RELATION SHIPS AMONGST THE CELLS) FOR STORAGE AND RETRIEVAL B BA BAC C D BC BAD BCC BCA BCB

QUAD-TREE STRUCTURE : CRITIQUE-2 AB CD A n n REQUIRES LESS STORAGE THAN RASTER AS NUMBER OF GRIDS WILL BE LESS GRID B WILL BE REPRESENTED BY ONLY 5 CELLS AGAINST 16 IN RASTER MODE B BA BAC C D BC BAD BCC BCA BCB

QUAD-TREE STRUCTURE - CRITIQUE-3 AB CD A n n n AMENABLE TO VARIABLE SCALE DATABASE HANDLING FACILITATES DISTRIBUTED DATABASE HANDLING OF LARGE DATABASES VARIOUS OTHER QUADTREE APPROCHES B BA BAC C D BC BAD BCC BCA BCB

QUAD-TREE STRUCTURE - CRITIQUE-4 n n VERY SENSITIVE FOR CO-ORDINATE TRANSFORMATION u TRANSLATION u ROTATION ENTIRE TREE STRUCTURE HAS TO BE RE-WORKED ON AFFECTING THESE TRANSFORMATIONS

QUAD-TREE STRUCTURE - CRITIQUE-5 n MANY OTHER APPROCHES PM QUADTREE u u n POINT QUADTREE u n SUB-DIVISION BASED ON DECOMPSITION OF BOUNDARY EDGES AND VERTICES ORIGINAL VECTOR STRUCTURE AND TOPOLOGY IS MAINTAINED, THUS GIVES BEST OF BOTH WORLD SUB-DIVISION BASED ON LOCATION OF ORDERED DATA POINTS RATHER THAN REGULAR SPATIAL DECOMPOSITION MX , PR, AND MANY MORE. . .

SPATIAL DATA MODELS - CRITIQUE n n n MODEL AFFECTS ALL ASPECTS OF GIS EFFICIENCY DIFFERENT APPROACHES STRIVE TO REALISE IDEAL REPRESENTATION OF FEATURES IN GIS VIS -A-VIS THE ASPECTS GIVEN IN TABLE NONE OF THE APPROACHES GIVES IDEAL SOLUTION

NON SPATIAL DATA HANDLING IN GIS n BASED ON DATABASE MANAGEMENT SYSTEM (DBMS) CONCEPTS

DATABASE MANAGEMENT SYSTEM (DBMS) DATABASE u STRUCTURED/ SYSTEMATIC, SHARABLE, NON-REDUNDANT STORAGE OF DATA DBMS u STANDARD S/W TOOL FACILITATING ALL ASPECTS OF DATABASE MANAGEMENT VIZ INPUT, STORAGE, RETRIEVAL & PRENENTATION

PURPOSE OF DBMS APPROACH n n n REDUCED DATA REDUNDANCY - VOLUME, CONSISTENCY, INTEGRITY, QUALITY DATA DICTIONARY - SELF DESCRIPTIVE, DOCUMENTED STORAGE PROGRAM & DATA INDEPENDENCE - INDEPENDANT CHANGE OF USER PROGRAM AND/OR DATA STORAGE FORMAT ( ITEMS, FILES, DEVICE) TRANSPARENCY - USER FREE FROM BOTHERATIONS OF INTERNAL STORAGE INTRICACIES MULTIPLE USER VIEWS CONCEPTUAL VIEW STORED DATABASE (INTERNAL VIEW) DBMS EXTERNAL VIEW- A EXTERNAL VIEW- B USER- A USER- B USER- C USER- D

DBMS TYPES (MODELS) - HEIRARCHIAL n RIGID VS HORIZONTAL RELATIONS OCCURANCES CAN NOT BE CHANGED COUNTRY STATE n REGION CITY TOWN

DBMS MODELS- NETWORK OCCURANCES CAN NOT BE CHANGED COUNTRY STATE n REGION CITY TOWN

DBMS MODELS- RELATIONAL n FLEXIBLE RELATIONSHIPS, OCCURANCES REDUNDANCY COUNTRY STATE C 1 S 1 C 2 S 2 C 3 S 3. . COUNTRY REGION C 3 R 1 C 3 R 2 C 2 R 3. . STATE n C 1 COUNTRY C 2 C 1 STATE S 1 S 3. . R 1 R 2 S 3 S 1 S 2 CITY C 1 C 2 C 3. . REGION TOWN. . . C 3 REGION C 3 STATE TOWN S 3 T 1 S 3 T 2. . T 1 T 2 TOWN REGION CITY. . STATE REGION. .

GIS-VS-DBMS TWO WAYS TO USE DBMS CONCEPT IN GIS n TOTAL DBMS SOLUTION - ALL DATA OPERATIONS IN DBMS ENVIRONMENT u PROBLEMS F F F n MIXED SOLUTION - GEO RELATIONAL u u n VARIABLE LENGTH OF RECORDS FOR CO-ORDINATES INFINITE NUMBER OF RELATIONSHIPS AMONGST THE SPATIAL OBJECTS INTERNAL STRUCTURE TOO COMPLEX FOR DBMS'S TO PROVIDE TRANSPARENCY SOME DATA ELEMENTS (ATTRIBUTES, ENTITY RELATIONSHIPS) IN DBMS (INFO, DBASE, ORACLE, INFORMIX) SOME DATA (LOCATIONAL) DIRECTLY FROM FILES BECAUSE DIFFICULT TO HANDLE IN DBMS MODEL SECOND METHOD GEO RELATIONAL USED IN GIS

GEO RELATIONAL MODEL IN GIS A 1 N 3 N 5 A 2 P 1 N 2 N 1 A 3 P 4 A 6 A 5 P 3 A 4 P 2 N 4 A 7 POLYGONS ARCS ID A 1 A 2 A 3 GEO DATA IN PROPRIETORY FORMAT - HIDDEN FROM USERS CO-ORD. . . ID P 1 P 2 P 3 ARCS A 1, A 2, A 3 A 2, A 7, A 5, A 6 A 3, A 5, A 4 ARCS ID L-POLY R-POLY FNODE TNODE LEN A 1 P 1 0 N 3 N 1. . . A 2 P 1 P 2 N 3. . . DESCR. . . . NON-SPATIAL DATA IN RELATIONAL (RDBMS) - ACCESSIBLE NODES ID N 1 N 2 N 3 CO-ORD X, Y POLYGONS ID AREA PERI DESCR POP. . P 1. . . . P 2. . . .