1. Application of GIS in hydrology and water resources management
University of Stuttgart – ENWAT
Part 1: July 24 – 26, 2007
Part 2: ?
Dietrich Schröder
University of Applied Sciences Stuttgart

2. Objective:
Almost everything that happens, happens somewhere.
80% of all decisions in policies are spatial related.
Geographical Information Systems are today state-of-the-art tools for supporting any kind of spatial related modelling and decisions.
This holds in particular in hydrology and water resource management.

3. Aims of the course (part 1):
Theoretical background of handling spatial related data
Basic principles of GIS
Handling of a GIS by example ArcGIS
in the context of water resource management
=> GIS is still (and will remain) an expert system!! (which is impossible to learn in a 2.5-day seminar)

4. Aims of the course (part 2): GIS in hydrology
DEM
GIS and surface hydrology
GIS and groundwater models
Case studies

5. Part 1: 24 July – 26 July 2007 Tuesday, 24th July Wedneday, 25th July:
Morning session: lectures
Afternoon session: exercises in the lab
Wedneday, 25th July:
Thursday, 26th July:
Examination: test

6. Outline Part 1
Overview of GIS
Data modeling
Spatial referencing: Coordinate systems, geodetic datum, and map projections
Georeferencing of images
Design of Thematic maps
Analysis of spatial data

7. What is Geographic Information System?
A GIS is a computer-based information system that enables
capture,
modeling, storage,
retrieval,
sharing,
manipulation,
analysis, and
presentation
of geographically referenced data

8. GIS and GIS Application
Don‘t mix it:
The software package (e.g. ArcGIS)
The application (e.g. hydrological analysis tools)
You can apply the same software product for many different applications (and of course use different software for the same application!)

9. Software: ArcGIS, GeoMedia, ILWIS, Smallworld, GRASS,...
Application: Forestry IS, DSS for urban planning, utility management system, hydrological analysis system,...

10. Development of GIS: from specialist system to Multi-Purpose GIS
utility management (network)
remote sensing (raster)
mapping (cartography)
...
environmental GIS (areal analysis)
multi-purpose GIS
Integrated
Multi-discipline approach

11. Components of GIS
Data Analysis
Hardware (computer, periphery like digitizing devices, high-end plotters)
Software (including generic procedurs for analysis)
Data
User (his or her expert knowledge
Data Visualization
Data Management

12. Life time of the components
Life time of the components H
hard-ware
computer, periphery
(digitizer, scanner, plotter)
~ 4 years S
soft-ware
programs, extensions, tools, methods
~ 7 years D
data
data, rules, knowledge
~ 25 years U
user

13. The GIS Cost Pyramide ~ 10 % Hard- ware ~ 15 % ~ 25 % Software ~ 50 %
Staff / Users
Data

14. Galery of GIS Applications
hydrology
cadastre
agriculture
environment
geo-sciences
forestry
emergency services
public services (utilities)
navigation
regional/local/rural/urban planning
tourism ...

15. Classical Application: Cadastre
Cadastral data is the geo base data for many large scale applications!

16. Cadastral data Germany:
ALK (Automatisierte Liegenschaftskarte)
Whole coverage of Germany
Scale level 1:1000 to 1:10 000
Accuracy some centimeter
Content: land parcel (with land use) and buildings

17. Flood inundation mapping for risk evaluation of building insurances

18. ZÜRS (Zonierungssystem für Überschwemmung, Rückstau und Starkregen)
Zoning system for inundation, backwater, and strong rain
1 river bed
2 Susceptibility class 3
(10 years)
3 Susceptibility class 2 (10 – 50 year)
4 Susceptibility class 1 (> 50 year)

19. Utilities: Sewerage Network

20. Utility Network and Orthophoto

21. Geo base data 1: 10 000 to 1:50 000 (ATKIS Authorative Topographic-Cartographic Information System)

22. ATKIS
Object-based structure
Whole coverage of Germany
Scale level 1: to 1:50 000
Accuracy 3 m (as TK25)
Polygons, lines, and points with attributes
Land use, river, transportation, etc.

23. Hydrography from ATKIS

24. Analysis of networks is a common task for GIS: rivers, utilities, transportation, …

25. Combining different maps
Land use
Soil
Geology
Elevation
Soil depth …
Result: HRU

26. Actual Trends
3D landscape visualization and animation
3D city models
Mobile GIS
Spatial enabled database systems
Web Services (SOA)

27. 3D landscape visualization
Vierwald-stätter See
e.g. ArcGis 9.3, G-Graphix (Freiburg), Skylinesoft (Israel)

28. 3D City Model Hamburg

29. 3D City Model Stuttgart

30. Mobile GIS

31. Tablet (rich client) to smart GPS receiver (thin client)

32. Spatial enabled database systems
In recent years more and more functionality has moved from GIS to database systems like Oracle (locator and spatial)

33. Oracle Spatial linear referencing system
Over 400 Spatial functions such as centroids and aggregate functions (e.g. unions
and user defined aggregates)
GeoRaster data type that natively manages georeferenced raster imagery
A data model to store and analyze network (graph) structure
A data model and schema to persistently store and update topology
Spatial analytic functions
• 3-dimensional data type support for terrain and city models and virtual worlds,
support for LIDAR-based map production
Spatial web services support (WFS 1.0, WFS-T 1.0, CSW 2.0, OpenLS 1.1, web

34. Web Services (Service oriented Architecture)
From desktop GIS to server GIS

35. GIS Classification Professional GIS (full functionality)
Desktop GIS (reduced functionality, mainly for visualization of spatial data)
CAD GIS (program with GIS functionality based on a CAD)
Internet-GIS (GIS Server) (Client-Server architecture with a Web-Browser and an application server)
Business-Map-GIS (simple cartographic tool with some GIS functionality)
Mobile-GIS (for mobile use, in particular for data collection and updating)

36. Top Ten 2004 per seats (professional GIS)
ArcGIS
ESRI
GeoMedia Professional
Intergraph
Erdas Imagine
Geosystems
GRASS
GRASS Developer team
30 000
Smallworld
GE Network Solution
Geomatica – EASI/Pace
CGI Systems
12 000

37. Top Ten 2004 per seats (desktop GIS)
ArcView 3.3
ESRI
GeoMedia
Intergraph
MapInfo Professional
MapInfo
MicroImages TNTAtlas + TNTlite
GIS Team
20 000
Sicad Spatial Desktop
Sicad Geomatics
10 000 …

38. Top Ten 2003 per seats (others)
Autodesk Map
Autodesk
(?)
CAD GIS
Geograf/Ingrada
HHK / Softplan
12 500
MicroStation Geographics
Bentley Systems
6700
Mappoint
Microsoft ?
Business GIS
Regio Graph
Macon AG
25 000 …

39. Summary GIS for visualization GIS for integration GIS for analysis
of spatial data
In both surface water and groundwater modeling, data management plays a major role. The compilation, analysis, and formulation of model input are the major phases of any modeling study, as is the creation of high-quality, graphical model output.
Much of the data required for model development, including land use maps, soil types, production well locations, basin delineation, water quality, recharge, evapotranspiration, parcel data, etc. are being made available in the form of GIS coverages. A GIS allows users to take advantage of the vast quantity of data available today for water resource applications.

40. Literature
M. Gurnell, D. R. Montgomery (Editors): Hydrological Applications of GIS (collection of papers)
J. Fürst: GIS in Hydrologie und Wasserwirtschaft, Wichmann (in German but with many references)
D.R. Maidment: ArcHydro, ESRI Press
D.R. Maidment: Hydrologic and Hydraulic Modeling Support with GIS (HEC-Ras related)
M.N. DeMers: Fundamentals of Geographic Information Systems, Wiley&Sons

41. Spatial Modeling

42. Outline
Modeling the real world
Discrete objects and continuous fields
Vector and raster representation
Managing spatial and attributive data
Object based data models
Process modeling
Dimensions
Topology

43. Modeling the real world
map
Modeling the real world
process model
real world
Perception of the world:
models of geographic phenomena
database
representation

44. What is our „model“ of the river Neckar?
A polyline on the a map 1: ?
A polygon on a map 1:1000?
A trench in the earth surface?
A habitat for animals and plants?
Is it a waterbody exactly defined by its channel banks?
Is it a continuous varying field of the river bed‘s elevation?

45. Moving from the real world through various data models to model output requires transformations in both information structure and information content. These transformations from the real world to binary representations include:
abstraction, generalization and selection of relevant concepts, processes and relationships in the real world
conceptual modeling of the relationships between abstract entities
mathematical modeling of the relationships between defined entities
physical sampling of the real world
storage of data in computers - may or may not include the necessity to model space
transformation of data between different representations (models).

46. Something of interest that Can be named or described
Geographic Phenomena
Something of interest that
Can be named or described
Can be georeferenced
(Can be assigned a time (interval) at which it is/ was present)
What is present? building, river,…
Where is it?  coordinates
(When was it present  time interval)

47. Types of geographic phenomena
A geographic field is a geographic phenomenon for which, for every point in the study area, a value can be determined.
Geographic objects populate the study area, and are usually well-distinguishable, discrete, bounded areas. The space between them is potentially empty.

48. Continuous Fields: elevation

49. Continuous Fields: rainfall

50. Continuous Fields: pH-value

51. Geographic Objects: Utility network
Sewage canal
Manhole
Valve
Building

52. Geographic Objects: river network

53. Discrete Fields: Land use classification

54. Discrete Fields: rainfall classes

55. Conceptual Representations
raster representation
model of geographic phenomena
real world
vector representation

56. Raster: For Continuous Fields
X Y Z

57. Raster representation
imaginary grid over the study area
only “inner” coordinates for each cell
whole grid has to be georeferenced
stores information on the interior of areal features
boundaries only implicit 1 . n
cell(35,67)
columns
rows n

58. location, value, resolution, and appearance
location: E, N
value: 24
resolution: 10m
appearance: RGB(255,0,0) 24 24

59. In GIS usually squares are used as cell form
Location of each cell as inner coordinates, specified are the coordinates of the upper left corner of the whole grid and ist direction towards north, e.g. for cell(35,45): x35=xul+35*resolution y45=yul-45*resolution
The meaning of the value depends on the application, e.g.
Intensity in a RADAR or LASER image
Classification of an RGB satellite image (range of values) for land use
Calculated combination of different band of a satellite image like NDVI (Normalized Difference Vegetation Index) using infra red
Elevation above sea level …

60. NDVI from NOAA-AVHRR satellite for February and May 1997 showing the photosynthetic activity

61. Floating point versus integer grids 1 2 3
Value land cover 1
forest 2
Farmland 3
grassland
VAT
Only integer grids can have a VAT (value attribute table)
Some operations only work on integer grids
Integer grids are mainly used for discrete fields like land use
Floating point grids are used for continuous fields like elevation (DEM)
grid

62. Raster Resolution

63. Cell raster: the value is an average value for the whole cell
Point raster: the value is the (interpolated) value of the cell center point

64. Advantages of the grid or raster representation
simple concept
easy management within the computer; many computer languages deal effectively with matrices, e.g. MatLab
map overlay and algebra is simple: cell by cell
native format of satellite imagery and scanned images
modeling and interpolating is simple, because the grid is dense and complete

65. Disadvantages of the grid or raster representation
fixed resolution, can’t be improved. So when combining maps of various resolutions, you must accept the coarsest resolution
information loss at any resolution, increasingly expensive storage and processing requirements to increase resolution
large amount of data especially at high resolution
not appropriate for high-quality cartography (line drawing)
slow transformations of projections (must transform each cell)
some kinds of map analysis (e.g. networks) is difficult or at least not ‘natural’
no correspondence to real world objects, i.e. difficult to link additional attributive data 

66. Rasterizing Continuous Fields

67. Rasterizing Discrete Objects

68. Cell size of the raster determines ist resolution: cell size max
Cell size of the raster determines ist resolution: cell size max. 50% of the smallest object to be recognized

69. Vector Data

70. Vector representation
point
line
Geometric primitives
area

71. Vector data and attributive data
Can be easily linked to additional information
Gage station:
Name, Code, temperature, discharge,…
River segment:
Name, Code, Length, quality, …

72. Advantages of the vector representation
precision is only limited by quality of the original data
very storage-efficient, since only points about which there is information or which form parts of lines and boundaries are stored
structuring the data logically is very easy
explicit topology makes some kinds of spatial analysis easy
high-quality output
Advantages of the vector representation

73. Disadvantages of the vector representation
not suitable for continuous surfaces such as scanned or remotely-sensed images and models based on these
time consuming capturing (digitizing, field survey)

74. Continuous fields in vector representation
TIN (Triangulated irregular network)

75. TIN as a Delaunay Triangulation
Non overlapping triangles, no other point in the outer circle of a triangle

76. Contour lines

77. Vectorizing rasters polygons lines
 Additional smoothing algorithm necessary!

78. Data Types in vector GIS
Geometry Data e.g / /
Graphical Description (symbology) e.g. width: 3mm color: blue pattern: dotted
Graphic Data: Geometry + Graphical Description X X X

79. Topology Information e.g. upID: 23 downID: 56
Attributive Data e.g. Type: ephemeral river length: 2 km Name: Thirsty Creek
Topology Information e.g. upID: downID: 56
Other Data e.g. Multimedia data ID
Type
Length_km
Name … 34
Ephemeral 2
Thirsty Creek

80. The feature representation of geographical objects
OID Geometry topology timestamp attributes
(neighboring polygons) landuse owner
12456 x1,y Jan forest Smith
xn,yn
A feature
has an object identity
has a property set (attributes)
usually has a geometry, i.e. one of the property is geometric associated with a reference system
may have various relationships to other features

81. Still one of the most often used format for data exchange in GIS!
How to store and manage this information?
Approach 1: separated storing of geometry, graphical description and attributive data
geometry
graphical description
(per feature class)
attributive data
OID
Still one of the most often used format for data exchange in GIS!
e.g. ESRI‘s ArcView Shape-Format:
geometry: Shape-File (*.shp)
attributive data: DBase-File (*.dbf)
„link“ or „index“ file (*.shx)
graphical description: Project-File (*.apr)

82. Most GIS support the direct import of CAD data
How to store ands manage this information?
Approach 2: storing geometry and the graphical description for each feature separated from the attributive data
geometry + graphical description
attributive data
OID
Most GIS support the direct import of CAD data
e.g. Autodesk‘s AutoCAD-Map:
geometry + graphical description: DWG or DXF-File
atributive data: Database-File

83. How to store ands manage this information?
Approach 3: storing geometry and attributive data together, but separated from the graphical description
geometry + attributive data
graphical description
feature class
e.g. ESRI‘s ArcGIS (Personal Geodatabase, real database systems like Oracle, Informix, DB2)
geometry + atributive data: Database-File (MS-Access, Oracle)
graphical description: Map-File (*.mxd)

84. Data Modeling in UML (Unified Modeling Language)

85. Features versus objects Only attributes, no operations
Mapping on „flat“ database tables structure like Personal Geodatabases
No inheritance
No complex objects, i.e. no nested tables

86. The model and its implementation

87. Process Modeling

88. Dimensions
2D:
without any height information, only latitude/longitude or Northing/Easting

89. 2D+1D:
position and contour lines two independent layers

90. 2.5D:
position with heights as attributive data

91. 3D surface:
to each position one height is associated
(sometimes also called 2.5 surfaces, don‘t get confused!)

92. 3D wire frame / volume:
to each position more than one height can be associated

93. 4th dimension: time
2+1+1D

94. time in ArcGIS

95. Topology
Structure of objects independent of their geometry. Using continuous transformation, topologic properties remain unchanged
“rubber sheet” transformation

96. Characteristics of the space in GIS
3D Euclidean space. Geometric primitives: points (zero-dimensional), lines (one-dimensional), areas (two-dimensional) and volumes (three-dimensional)
the space is a metric space, (distance function)
the space is a topological space
interior and boundary are invariant under topological mappings

97. Problems not using topology
lines between adjacent polygons must be digitized and stored twice (slivers and gaps)
there is no neighborhood information
islands are impossible except as purely graphical constructions
there is no easy way to check if the boundary is correct and complete (dead-ends, weird polygons)
slivers
dead ends
gaps
weird polygons

98. Vector Data Structures & Topology

99. Geographical Objects (Simple Features)
coordinates P1
x1,y1,x2,y2,x3,y3,…,x1,y1 P2
x1,y1,x9,y9,x8,y8,…x1,y1 …
Area-Object by area-object exist explicitly!
Topology only exist implicitly!

100. Interoperability

101. Simple Features – Geometry Definition
0, 1 and 2 - dimensional geometry
Object (Feature)
Point
Line / Polyline
Area (with holes)
Exterior
Border
Interior
Single objects
Groups of similar objects
Definitions for

102. Simple Features – Geometry Definition
Pecularities of simple feature geometries:
only straight connection of points (no curved arcs)
no topology
Examples for not allowed objects (not simple)

103. Topological Structure
node-arc-area (NAA) representation
arc
From
node
To node
Left
Right 1 2 P1 3 … 7 8 P2 10 13
node X y 1 X1 Y1 2 X2 y2 3 x3 y3 …
area
arc P1
1,2,3,4,5,6,7,8,9 P2
10,11,12,13,14,7,8,9 …
Area-object by area-object and topology exist explicitly!

104. Topology in ArcGIS
Topology has historically been viewed as a spatial data structure used primarily to ensure that the associated data forms a consistent and clean topological fabric. With advances in object-oriented GIS development, an alternative view of topology has evolved.
The geodatabase supports an approach to modeling geography that integrates the behavior of different feature types and supports different types of key relationships. In this context, topology is a collection of rules and relationships that, coupled with a set of editing tools and techniques, enables the geodatabase to more accurately model geometric relationships found in the world.

105. …

106. …

107. Maintaining data consistency & topology
some GIS maintain consistency geometrically, i.e. by coinciding coordinates
maintaining explicitly stored topology is very time consuming

108. no real standard for topology no real standard for symbology
Data exchange
de-facto-standards like Shape-Format or DXF (only geometry) based on simple features
no real standard for topology
no real standard for symbology
Metadata very important
OGC (Open Geospatial Consortium: initaitives on specification for interoperability
Simple feature specification
WMS and WFS specification
Symbology encoding
Geo Processing
tools for data transformation like FME (Feature Manipulation Engine) from Safe software

109. Summary
data models and their implementation should represent spatial phenomena completely and support efficient analysis
grids for continuous fields (and discrete fields)
discrete primitives for objects (and discrete fields)
vector-GIS organize geographic objects in feature classes
GIS are more or less still 2D and 3D surfaces
simple feature is the standard for data exchange
Topology helps maintaining consistent data sets
with WebServices data exchange problems will become less important