Geographical Information System

Definition. A Geographic Information System (GIS) is a computer-based system for capturing, storing, managing, analysing, and displaying data that is linked to locations on the Earth's surface. It combines spatial data (where things are) with attribute data (what they are like) so that questions about location and patterns can be answered and shown on maps.

Benefits of GIS

• Better decision making: Combines many layers of information so planners can compare options spatially.
• Efficient data management: Large volumes of geographic data are stored in one organized database, which is easy to update and retrieve.
• Powerful spatial analysis: Overlay, buffering, network analysis and terrain analysis give results that are hard to get manually.
• Improved communication: Maps and visual displays make findings easy for non-experts to understand.
• Saves time and cost: Automates map production and analysis, reducing repeated field work.
• Better planning and monitoring: Changes over time (land use, forest cover, urban growth) can be tracked and predicted.
• Disaster management: Hazard and risk mapping, and response planning.
• Data sharing and integration: Data from different sources and departments can be integrated and shared.

A map is a scaled graphic representation of the Earth's surface. The main elements of a map are:

1. Title: Tells what area and theme the map shows, for example "Land Use of Kathmandu District".
2. Legend (key): Explains the symbols, colours and patterns used on the map.
3. Scale: Relates map distance to ground distance. It may be a representative fraction (1:50,000), a verbal statement, or a graphic scale bar.
4. North arrow (orientation): Shows the direction of north so that the map can be oriented correctly.
5. Map body (data frame): The main area where the geographic features are drawn.
6. Grid, graticule and coordinates: Lines of latitude/longitude or projected coordinates that help locate positions.
7. Projection and datum information: States the coordinate system used, for example UTM Zone 45N, WGS84.
8. Border (neatline) and inset/locator map: The neatline frames the map; an inset shows the area's location in a wider region.
9. Source, author, and date: Gives the data source, producer, and year, showing the reliability of the map.

Remote sensing is the science and technique of acquiring information about objects or areas on the Earth's surface without direct physical contact, using sensors mounted on platforms such as satellites, aircraft, or drones that record electromagnetic radiation reflected or emitted from the surface.

How remote sensing works

1. Energy source or illumination (A): The Sun (passive remote sensing) or the sensor itself, such as radar or LiDAR (active remote sensing), provides electromagnetic energy.
2. Interaction with the atmosphere (B): The energy passes through the atmosphere and is partly absorbed or scattered.
3. Interaction with the target (C): The energy hits surface features and is absorbed, transmitted, or reflected depending on their properties. Different features (water, vegetation, soil, built-up areas) have different spectral signatures.
4. Recording by the sensor (D): The reflected or emitted energy is detected and recorded by the sensor on the satellite or aircraft.
5. Transmission, reception, and processing (E): The data is sent to a ground station, where it is corrected and turned into digital images.
6. Interpretation and analysis (F): Images are interpreted visually or by computer (classification) to extract information.
7. Application (G): The information is used for mapping, monitoring, and decision making.

Digital Terrain Model (DTM). A DTM is a digital representation of the bare-earth surface of the terrain, stored as a set of elevation values at regular grid points (raster) or irregular points (TIN) with their x, y, z coordinates. It excludes vegetation and buildings; a Digital Elevation Model (DEM) is often used for the same concept, while a Digital Surface Model (DSM) includes objects on the surface. A DTM can be created from contour lines, survey points, photogrammetry, LiDAR, or radar (SRTM).

Applications

• Slope, aspect, and terrain analysis: Used to derive slope maps, aspect maps, and landforms.
• Watershed and hydrological modelling: Delineating catchments, flow direction, drainage networks, and flood simulation.
• Hazard mapping: Landslide susceptibility and flood-risk zoning.
• Engineering: Route selection for roads and canals, dam sites, and cut-and-fill volume calculations.
• Visibility and viewshed analysis: Placement of telecom towers and viewpoints.
• 3D visualization and orthophoto production: Perspective views, flight simulation, and correction of satellite images.
• Agriculture and forestry: Land suitability and terrain-based planning.

A GIS is made up of functional components that carry out the sequence of operations from getting data to producing output.

1. Data input (capture): Converting data into digital form from sources such as digitizing and scanning maps, GPS surveys, remote-sensing images, field surveys, and existing databases. Includes editing and error checking.
2. Data storage and management: Storing spatial and attribute data in a structured database (files or a DBMS/geodatabase), with organization, indexing, retrieval, and updating of data.
3. Data manipulation and transformation: Preparing data for analysis, including map projection change, coordinate transformation, edge matching, generalization, and format conversion.
4. Data analysis: The core of GIS, which includes query, overlay, buffering, network analysis, interpolation, and terrain modelling to answer spatial questions.
5. Data output (visualization): Presenting results as maps, tables, charts, reports, 3D views, or digital files.

Alongside these functions, a complete GIS also needs hardware, software, data, people, and methods as its supporting components.

A good database design makes the data reliable, efficient to use, and easy to maintain. Its main characteristics are:

• Minimal redundancy: Each fact is stored once (normalization), which saves space and avoids update anomalies.
• Data integrity and accuracy: Constraints such as primary keys, foreign keys, and validation rules keep the data correct and consistent. In a spatial database this includes topological integrity.
• Consistency: The same data is the same everywhere in the database.
• Data independence: Changes in the physical storage do not affect the application programs.
• Efficient retrieval and performance: Indexing (including spatial indexes) allows fast queries.
• Security and access control: Only authorized users can read or change data; backup and recovery are provided.
• Flexibility and scalability: The design can accommodate new data, users, and growth.
• Ease of maintenance and updating: A clear structure, documentation, and standard naming keep it easy to update.
• Data sharing and integration: Many users and applications can use the data at the same time.
• Completeness: It meets all the information needs of the users.

(a) Map layers

A map layer is a collection of geographic features of the same theme and type, such as roads, rivers, land use, or contours, stored as a separate dataset. In GIS, the real world is represented by stacking many thematic layers over the same coordinate system, so each layer can be switched on or off, symbolized, and analysed independently. Layers can be combined using overlay operations to find relationships, for example overlaying slope, soil, and land use to select suitable farmland. Each layer has its own geometry (point, line, polygon, or raster) and attribute table. Layered organization makes data management and updating simple.

(b) Map projection

A map projection is a mathematical method for transforming the curved surface of the Earth (latitude and longitude) onto a flat plane (x, y coordinates). Because a sphere cannot be flattened without distortion, every projection distorts one or more of shape, area, distance, or direction.

• By property: conformal (preserves shape, e.g. Mercator), equal-area (preserves area, e.g. Albers), equidistant, and azimuthal.
• By developable surface: cylindrical, conical, and planar.

Examples: UTM (Universal Transverse Mercator, Nepal lies in zones 44, 45, and 46) and the Modified Transverse Mercator used by the Survey Department of Nepal. The choice of projection depends on the area and purpose of the map.

(c) Geodatabase

A geodatabase is a database, usually built on a relational DBMS, that stores spatial data (feature classes, rasters, topologies, networks) together with attribute tables, relationships and rules in a single container. It supports feature datasets, subtypes, domains, topology rules, relationship classes, and multi-user editing with versioning. Its advantages are centralized storage, data integrity, scalability, and support for complex behaviour. Types include personal, file, and enterprise (multi-user) geodatabases.

Raster data. Raster data represents geographic space as a regular grid of equal-sized cells (pixels). Each cell has a single value representing the attribute at that location, such as elevation, land-cover class, or reflectance. Examples are satellite images, scanned maps, and DEMs. The cell size defines the spatial resolution.

Vector data. Vector data represents features using points, lines, and polygons defined by x,y coordinates. A point stores a location (well, school), a line is an ordered set of points (road, river), and a polygon is a closed area (lake, district boundary). Attributes are stored in linked tables. Examples are shapefiles and CAD data.

Differences

Advantages and disadvantages

Raster

• Advantages: simple data structure; easy and fast overlay and map algebra; suits continuous data; direct use of satellite images; good for modelling and image processing.
• Disadvantages: large file sizes; resolution limited by cell size; boundaries look jagged; poor for network analysis; projection transformation may degrade data.

Vector

• Advantages: compact storage; accurate positions and sharp boundaries; topology is possible; good for network analysis; many attributes per feature; high quality cartographic output.
• Disadvantages: more complex data structure; overlay is slower and harder; poor for continuous surfaces; requires more processing for simulation and analysis; costly to capture by digitizing.

Spatial database design is the process of organizing geographic data (features, their locations and attributes, and relationships) into a structured database that supports storage, query, and analysis. It follows the standard database design stages with spatial additions.

Stages

1. Requirement analysis: Identify users, the purpose of the system, the questions to be answered, and the data needed.
2. Conceptual design: Define the real-world entities, attributes, and relationships (including spatial relations such as contains, adjacent to, intersects) using an ER model or UML. It is independent of software.
3. Logical design: Convert the conceptual model into the data model of the chosen system. Choose vector or raster, define feature classes (point, line, polygon), tables, primary and foreign keys, normalization, topology rules, and coordinate system and projection.
4. Physical design: Define storage structures, data types, spatial indexes (R-tree, grid), file formats or DBMS, access rights, and backup.
5. Implementation, testing, and maintenance: Create the database, load and validate data, test queries, and maintain and update it.

Example: Municipal ward and land-parcel database

Entities and layers:

• Ward (polygon): ward_id (PK), ward_no, area, population
• Parcel (polygon): parcel_id (PK), ward_id (FK), owner_id (FK), land_use, area
• Owner (table): owner_id (PK), name, citizenship_no
• Road (line): road_id (PK), name, type, width
• School (point): school_id (PK), name, ward_id (FK), students

Relationships: one Ward contains many Parcels (1:N); one Owner may own many Parcels (1:N); Roads pass through (intersect) Wards (M:N, solved by spatial overlay). All layers use one coordinate system (UTM Zone 45N), a spatial index is created on each geometry column, and topology rules such as "parcels must not overlap" and "parcels must lie within a ward" ensure integrity.

Sample query: Find all parcels in ward 5 within 100 m of a road: select parcels where ward_id = 5 and the parcel lies within the 100 m buffer of the Road layer. This shows how attribute and spatial data work together in one design.

Nepal is a mountainous and developing country with steep terrain, scattered settlements, limited resources, frequent natural disasters (earthquakes, floods, landslides), and a need for planned development after the shift to federalism. GIS helps by integrating data for evidence-based planning at low cost, and it plays a role in many sectors.

Roles of GIS in Nepal

• Disaster risk management: Mapping hazard-prone areas for flood, landslide, earthquake, and glacial lake outburst, and planning evacuation routes and relief. After the 2015 Gorkha earthquake, GIS was used for damage assessment and reconstruction planning.
• Agriculture and food security: Land-suitability analysis, crop monitoring using satellite images, and irrigation planning.
• Forestry and environment: Community forest management, forest cover change, wildlife habitat (national parks), and watershed management.
• Infrastructure and energy: Selecting routes for roads, transmission lines, and hydropower sites using terrain data.
• Land administration: Digital cadastral mapping and land records in the Land Management and Survey Departments.
• Local government planning: Municipal and ward-level planning, property tax mapping, and service location under federalism.
• Health and education: Locating facilities, planning new schools or health posts, and disease mapping.
• Tourism and climate studies: Trekking route maps and monitoring of glaciers and snow cover.
• Census and elections: Constituency delimitation and spatial presentation of census data.

Application: landslide hazard mapping

Landslides cause many deaths and losses in Nepal every monsoon. A GIS-based hazard map is prepared as follows:

1. Collect layers: slope and aspect (derived from a DEM), geology, land cover, rainfall, distance to rivers and roads, and the inventory of past landslides.
2. Reclassify each layer into ranks and assign weights according to how much it contributes to landslides.
3. Combine the layers with a weighted overlay to obtain a landslide susceptibility score for each cell.
4. Classify the result into low, moderate, high, and very high hazard zones and produce the map.

Government agencies such as the Department of Mines and Geology and local governments use these maps to avoid building settlements and roads in high-risk zones, to decide relocation of vulnerable households, and to plan mitigation measures. The result is safer and cheaper planning than field survey alone, which would be difficult across Nepal's rugged terrain.

GIS is a computer-based system for capturing, storing, managing, analysing, and presenting spatial data linked to locations on the Earth. Its main components are hardware, software, data, people, and methods.

Functions

• Data capture/input: Digitising, scanning, GPS, and remote sensing.
• Data storage and management: Spatial and attribute data in a database.
• Data manipulation: Projection change, format conversion, and editing.
• Spatial analysis: Query, overlay, buffering, network analysis, and surface analysis.
• Output/visualisation: Maps, reports, charts, and 3D views.

Applications

• Natural resource management (forests, water, soil).
• Disaster management and risk mapping.
• Urban and regional planning, land administration.
• Agriculture and precision farming.
• Transportation and logistics routing.
• Public health and disease mapping.
• Utilities (water, electricity, telecom) and business site selection.

Map projection is a systematic mathematical transformation of positions on the curved Earth (latitude and longitude) to a flat plane (x, y). Distortion in shape, area, distance, or direction is unavoidable.

Techniques (by developable surface)

1. Cylindrical projections: The globe is projected onto a cylinder wrapped around it, which is then unrolled. Examples are Mercator (conformal) and Transverse Mercator/UTM. They suit world maps and, in transverse form, narrow north-south zones.
2. Conic projections: A cone is placed over the globe, usually touching along a standard parallel. Examples are Lambert Conformal Conic and Albers Equal-Area. They suit mid-latitude regions with east-west extent.
3. Azimuthal (planar) projections: A plane touches the globe at one point, usually a pole. Examples are Stereographic and Orthographic. They suit polar regions and air-route maps.

Classification by property preserved

• Conformal: Preserves shape and angles locally (Mercator).
• Equal-area: Preserves area (Albers, Mollweide).
• Equidistant: Preserves distance from certain points or lines.
• Compromise: Balances distortions (Robinson).

The choice depends on the area, purpose, and the property that matters most.

GPS (Global Positioning System) is a satellite-based navigation system owned by the USA that gives three-dimensional position (latitude, longitude, height) and time anywhere on Earth, in any weather. A receiver measures the distance to at least four satellites by the travel time of radio signals (trilateration) and solves for its position and clock error.

Components

1. Space segment: A constellation of about 24 or more satellites in six orbital planes at about 20,200 km altitude. Each satellite carries atomic clocks and transmits signals with its position and time.
2. Control segment: Ground stations (a master control station, monitor stations, and ground antennas) that track the satellites, compute orbits and clock corrections, and upload them.
3. User segment: The receivers and users, such as handheld units, phones, survey-grade receivers, and vehicle units, which decode the signals and compute position.

Other constellations (GLONASS, Galileo, BeiDou) work on the same principle.

Grid-like (raster) format

Space is divided into a regular matrix of cells arranged in rows and columns. Each cell stores one value for a theme (elevation, land cover, reflectance). Location is implied by the row and column position, and the cell size sets the resolution. Typical examples are satellite images and DEMs. Advantages are a simple structure, easy overlay (map algebra), and suitability for continuous surfaces. Disadvantages are large file size, limited precision, and no explicit topology. Storage can be reduced by compression such as run-length encoding and quadtrees.

Coordinate (vector) format

Features are stored as explicit coordinates (x, y) of vertices. A point is a single coordinate pair, a line is an ordered series of points, and a polygon is a closed line. Attributes are linked to each feature, and topology can record connectivity and adjacency (arc-node structure). Advantages are compact storage, accurate boundaries, scale-independent display, and network analysis. Disadvantages are a complex structure, slower overlay, and weak representation of continuous surfaces.

Both can be converted into one another (rasterisation and vectorisation).

A geodetic datum defines the size and shape of the Earth model (ellipsoid) and how it is positioned relative to the Earth. The International Terrestrial Reference Frame (ITRF) is a geocentric, globally defined frame whose origin is at the Earth's centre of mass, realised by a network of stations measured with GNSS, VLBI, SLR, and DORIS.

Benefits of an ITRF-based datum

• Global consistency: Coordinates are compatible worldwide, so national data can be integrated with other countries.
• Compatible with GPS/GNSS: GPS (WGS84) is aligned with ITRF to the centimetre level, so measurements need no complex transformation.
• High accuracy: Positions are precise to centimetre or millimetre level.
• Accounts for crustal motion: Station velocities are given, which is important in a tectonically active region like the Himalayas; coordinates are tied to an epoch.
• Reliable monitoring: Supports sea level, plate movement, and deformation studies after earthquakes.
• Reduced mismatch: Avoids offsets between old local datums and modern satellite positioning.
• Suits modern applications: Cadastre, engineering, navigation, and GIS data sharing.

Agriculture is the backbone of Nepal's economy, and GIS supports it in these ways.

• Land suitability and crop planning: Overlaying soil, slope, rainfall, temperature, and elevation to find the best crops and areas.
• Crop monitoring and yield estimation: Satellite images (NDVI) track crop health, area, and expected yield.
• Precision farming: GPS and GIS guide inputs such as fertiliser and water to where they are needed.
• Irrigation planning: Mapping water sources, canals, and command areas.
• Soil and land mapping: Identifying degraded, erosion-prone, or fertile land.
• Disaster and pest management: Mapping drought, flood, and pest outbreaks for early warning and relief.
• Market and supply chain: Locating roads, markets, and collection centres to reduce transport costs.
• Policy and statistics: Land-use planning and subsidy targeting by local governments.

This is valuable in Nepal's fragmented, hilly terrain, where field surveys are costly.

(a) Map features

Map features are the real-world objects and phenomena shown on a map. They are represented as points (wells, schools), lines (roads, rivers, boundaries), and polygons (lakes, forests, districts). They can also be natural or cultural (man-made) features. Each feature has a spatial component (location and shape) and descriptive attributes, and it is shown with symbols, colours, and labels that are explained in the legend. Features are grouped into themes or layers.

(b) Surface modeling

Surface modelling represents a continuous phenomenon, such as elevation, rainfall, or temperature, whose value varies across space. It uses sample points, contours, or a grid to build a surface model, either a raster DEM or a TIN, with interpolation (IDW, kriging, spline) filling gaps. From the surface, slope, aspect, contours, volumes, viewsheds, and 3D views are derived. It is used in terrain analysis, hydrology, and engineering.

(c) Spatial database

A spatial database is a database optimised to store, index, and query data describing locations and shapes, along with their attributes. It supports geometry types (point, line, polygon), spatial operators (intersects, within, distance), spatial indexes (R-tree), and topology. Examples are PostGIS, Oracle Spatial, and ArcGIS geodatabases. It provides multi-user access, integrity, and security for GIS data.

"Data is fuel to GIS"

A GIS is only as useful as the data in it. Software, hardware, and people are like an engine, but without accurate, current, and relevant data the system produces nothing useful. Data collection is usually the most expensive and time-consuming part of a GIS project, so data drives the quality of every analysis and decision.

Types of geographic data

1. Spatial (geometric) data: Location and shape, stored as vector (points, lines, polygons) or raster (grids, images).
2. Attribute (non-spatial) data: Descriptive facts such as names, population, land-use class, kept in tables.
3. Temporal data: The time dimension of features.
4. Metadata: Data about data, including source, accuracy, projection, and date.

Sources of geographic data

• Existing maps (topographic sheets, cadastral maps).
• Field surveys (total station, levelling).
• GPS/GNSS measurements.
• Remote sensing (satellite images, aerial photographs, LiDAR, drones).
• Census and statistical records and government agencies.
• Existing digital data, open data (OpenStreetMap), and web services.

Data capturing methods

• Digitising: Tracing features from a paper map with a digitiser table (manual) or on screen (heads-up).
• Scanning: Converting a map to a raster image, followed by vectorisation.
• Field survey and GPS: Direct measurement of coordinates.
• Remote sensing and photogrammetry: Interpretation of images to extract features and elevation.
• Keyboard/tabular entry: Typing attribute data, or importing spreadsheets and databases.
• Data transfer: Converting from other digital sources and formats.
• Crowdsourcing/mobile apps: Collection by volunteers or field teams.

Captured data must be checked, edited, and georeferenced before use.

Geodatabase

A geodatabase is a database in which spatial data (feature classes, rasters, networks, topologies) and attribute tables are stored together in a relational DBMS or file system. It supports behaviours such as domains, subtypes, relationship classes, topology rules, and versioning for multi-user editing. It is the native format of ArcGIS and can be personal, file, or enterprise (using SQL Server, Oracle, or PostgreSQL).

Steps for geodatabase design

1. Identify the information products and requirements: Decide what maps, reports, and analyses users need.
2. Identify the map layers and data: List the themes (roads, rivers, parcels) and the geographic representation of each (point, line, polygon, raster), scale, and accuracy.
3. Define attributes: For each layer, specify fields, data types, and domains.
4. Group the layers: Organise related feature classes into feature datasets sharing one coordinate system.
5. Define relationships and rules: Create relationship classes, topology rules (for example, no overlaps), and subtypes.
6. Create a conceptual and logical model: Draw a diagram (UML or ER) of classes and relationships.
7. Implement: Build the schema in the software, set the spatial reference and indexes.
8. Load data and test: Import data, check quality, and run typical queries.
9. Maintain and document: Set up backup, security, versioning, and metadata.

Hydrology modeling

Hydrology modelling is the use of mathematical and computer models to simulate the movement, distribution, and quantity of water in a watershed, covering rainfall, runoff, infiltration, evaporation, streamflow, and floods. Models predict how much water flows, where, and when.

Role of GIS in hydrology

GIS provides the spatial data and tools for a model and displays the results.

• Terrain analysis from a DEM: Fill sinks, flow direction, flow accumulation, stream networks, and watershed delineation.
• Preparing model inputs: Soil type, land cover, slope, and rainfall layers are overlaid to derive parameters such as the Curve Number for runoff.
• Flood mapping: Flood inundation zones are modelled and mapped for risk and warning, using tools such as HEC-HMS and HEC-RAS linked to GIS.
• Water resource management: Locating suitable sites for reservoirs, hydropower, and irrigation; groundwater recharge zones.
• Water quality and pollution: Mapping sources and spread of contamination.
• Climate and land-use change: Simulating effects on runoff.

Challenges GIS helps solve: Flooding, water scarcity, erosion and sedimentation, and poor data coverage. In Nepal, GIS-based hydrologic models are valuable for steep, data-scarce Himalayan basins and for flash flood early warning.

Main components of a GIS

1. Hardware: Computers, servers, GPS units, scanners, digitisers, and printers/plotters. It provides the processing, storage, and input/output capability on which the system runs.
2. Software: GIS software such as ArcGIS or QGIS and the DBMS, with tools for input, storage, analysis, and display. It provides the functions for editing, analysis, and map-making.
3. Data: Spatial data (vector and raster) and attribute data. It is the most valuable part, as the quality of data determines the quality of results.
4. People: Users, analysts, managers, and developers who design, operate, and interpret the system. They turn the tools into useful solutions.
5. Methods (procedures): Rules, standards, workflows, and organisational practices. They ensure consistent, accurate, and repeatable results.

Contribution: Hardware runs the software, which processes the data under the guidance of people following defined methods. If one component is weak, the whole GIS suffers.

Types of map projections

1. Cylindrical (for example, Mercator, UTM)

• Advantages: Simple grid; north is always up; Mercator is conformal and good for navigation; UTM is accurate for narrow zones.
• Disadvantages: Large area distortion away from the standard lines, especially near the poles (Greenland looks as big as Africa).

2. Conic (for example, Lambert Conformal Conic, Albers)

• Advantages: Low distortion along standard parallels; good for mid-latitude countries with east-west extent; can be conformal or equal-area.
• Disadvantages: Distortion increases away from standard parallels; unsuitable for the whole world or polar areas.

3. Azimuthal/planar (for example, Stereographic, Orthographic)

• Advantages: Directions from the centre are true; good for polar regions and air routes.
• Disadvantages: Distortion increases away from the centre; shows only part of the globe.

By property

• Conformal: preserves shape, but distorts area.
• Equal-area: preserves area, but distorts shape.
• Equidistant: preserves distances only along certain lines.

No projection preserves everything, so the choice depends on the purpose and extent.

Principles of GIS database design

• Understand user requirements: Design starts from the needed outputs, queries, and users.
• Stepwise design: Conceptual, logical, then physical design.
• Layered organisation: Group data by theme into layers or feature datasets with a common coordinate system.
• Integrity and consistency: Use keys, domains, and topology rules to keep data valid.
• Minimal redundancy: Normalise tables and store each fact once.
• Standards and metadata: Use agreed formats, naming, and documentation.

Key considerations for efficiency

• Appropriate data model: Choose vector or raster to suit the data and scale.
• Spatial indexing: R-tree or grid indexes speed up queries.
• Accuracy, scale, and resolution: Match the data detail to the purpose.
• Common spatial reference: One projection and datum avoids misalignment.
• Scalability and performance: Plan for growth in data and users, with tiling and partitioning where needed.
• Security and backup: Control access and maintain recovery plans.
• Interoperability and maintenance: Support data exchange and ease of updating.

Remote sensing is a major source of data for GIS, supplying up-to-date information over large and inaccessible areas.

Roles in data capture

• Land use/land cover mapping: Image classification produces thematic layers for GIS.
• Base map creation and updating: Roads, buildings, rivers, and settlements are extracted from high-resolution images and drone photos.
• Elevation data: Stereo imagery, radar (SRTM), and LiDAR produce DEMs.
• Change detection: Comparing images of different dates shows deforestation, urban growth, or glacier retreat.
• Vegetation and resource monitoring: Indices such as NDVI show crop and forest health.
• Disaster mapping: Quick mapping of floods, landslides, and earthquake damage.
• Georeferenced raster backdrop: Images form a visual base layer for digitising.

Benefits: Wide coverage, repeated observations, low cost per area, and digital format ready for GIS. Limits: Cloud cover, limited resolution, and the need for ground verification and processing.

GPS positions are affected by several errors.

1. Satellite clock errors: Small timing errors in satellite clocks cause range errors.
2. Orbital (ephemeris) errors: The broadcast satellite position differs slightly from the true position.
3. Ionospheric delay: Charged particles slow and bend the signal; this is a large error source, especially in solar activity.
4. Tropospheric delay: Water vapour and pressure in the lower atmosphere delay the signal.
5. Multipath error: The signal reflects off buildings, water, or terrain before reaching the receiver, so the path appears longer.
6. Receiver errors: Receiver clock bias, noise, and antenna phase-centre errors.
7. Satellite geometry (DOP): Poor geometry, when satellites are clustered together, increases the position error. A low PDOP is better.
8. Selective availability (historical): Intentional degradation, switched off in 2000.
9. Signal obstruction and user errors: Dense canopy, deep valleys, and urban canyons block signals; wrong datum or antenna height also affects results.

Corrections: Differential GPS (DGPS), RTK, dual-frequency receivers, and post-processing reduce these errors.

Surface modeling represents a continuous phenomenon, where each location has a value (elevation, rainfall, temperature), as a surface that can be analysed and displayed. In GIS it is built from sample points, contour lines, or spot heights.

Methods

• Raster DEM (grid): Elevation values on a regular grid.
• TIN: Irregular points connected into non-overlapping triangles that follow terrain breaks.
• Interpolation: IDW, spline, and kriging estimate values at unsampled locations.

Representing terrain features

• Contours and hillshade show relief.
• Slope and aspect show steepness and direction.
• Drainage: Flow direction, streams, and watersheds.
• Ridges, peaks, valleys, and pits can be extracted.
• Viewshed shows visible areas from a point.
• 3D perspective views and volume (cut and fill) calculation.

Surface models help in engineering, hazard mapping, hydrology, and planning.

(a) Future of GIS

GIS is moving toward cloud-based and web GIS, real-time data from IoT sensors and mobile devices, and integration with artificial intelligence and machine learning for automated feature extraction and prediction. Other trends are 3D and digital twins, drone and high-resolution satellite data, big data analytics, augmented reality, open-source tools, and wider use in smart cities, climate adaptation, and everyday mobile applications. GIS will become more accessible and more embedded in routine decision-making.

(b) Geodatabase

A geodatabase stores spatial data (feature classes, rasters, topologies, networks) and attribute tables together in a relational database. It supports domains, subtypes, relationship classes, topology rules, and multi-user versioned editing. Types include personal, file, and enterprise geodatabases. It gives better data integrity, security, and performance than shapefile collections.

(c) Elements of map

The main elements are the title, map body, legend, scale, north arrow, neatline, grid or graticule, locator inset, and source/credit notes. Together they make a map readable, accurate, and attractive.

Overview

• Vector: Features as points, lines, and polygons defined by coordinates, with attributes in tables.
• Raster: A grid of equal-sized cells, each with a value.
• TIN (Triangulated Irregular Network): A vector-based surface model made of non-overlapping triangles built from irregularly spaced points, each with an x, y, z value.

Comparison

Advantages and disadvantages

• Vector: Accurate boundaries, small files, good cartographic output, supports networks. Disadvantages are complex structure, slow overlay, and weak for continuous data.
• Raster: Simple, fast analysis, good for images and surfaces. Disadvantages are large files, limited precision, and blocky output.
• TIN: Efficient storage in varied terrain, preserves breaklines such as ridges and streams, and gives accurate surfaces. Disadvantages are costly construction, difficult to handle for large areas, and surfaces need specialised software.

Typical applications

• Vector: Cadastre, road and utility networks, administrative boundaries, site selection.
• Raster: Satellite image analysis, land-cover maps, DEMs, suitability and hydrological modelling.
• TIN: Terrain modelling, slope and volume calculation, road and dam design, 3D visualisation.

Principles of vector-based spatial analysis

Vector analysis uses the exact geometry of points, lines, and polygons, together with their attributes and topological relations (connectivity, adjacency, containment). Its principles are that features are discrete objects, that relationships such as "within", "intersects", and "touches" can be tested, and that new features and attributes can be derived from existing ones.

Techniques

1. Attribute and spatial query: Select features by value (population > 5000) or by location (inside a district).
2. Buffering: Create a zone at a given distance around features (for example, 100 m around rivers).
3. Overlay: Combine layers using union, intersect, erase, and identity to create new features with merged attributes.
4. Clip and dissolve: Cut one layer by another, and merge adjacent polygons with the same attribute.
5. Proximity analysis: Nearest feature and distance calculations.
6. Network analysis: Shortest path, service area, and routing along lines.
7. Measurement: Length, area, and perimeter.
8. Geocoding and spatial join: Link tables to locations and transfer attributes by location.

Examples

• Flood hazard: Buffer rivers by 200 m and intersect with a land-use layer to find houses at risk.
• School site selection: Select parcels outside a 500 m buffer around existing schools and within 100 m of roads.
• Emergency routing: Network analysis finds the shortest route from a fire station to an incident.
• Land-use change: Union of land-use layers from two years yields the change areas.
• Health service coverage: Buffers around health posts identify villages without access within 5 km.

Role of hydrology modeling in GIS

Hydrology modelling simulates rainfall, runoff, infiltration, and streamflow in a watershed. GIS supplies the spatial inputs and tools to build, run, and display these models.

• Terrain processing from DEM: Sink filling, flow direction, flow accumulation, stream network and watershed delineation.
• Parameter derivation: Overlaying soil, land cover, slope, and rainfall layers to obtain curve numbers and runoff coefficients.
• Simulation: Linking with models such as HEC-HMS, HEC-RAS, and SWAT to estimate discharge, flood depth, and extent.
• Visualisation: Maps of watersheds, flow paths, and inundation zones.

Application in water resource management

• Flood forecasting, flood hazard and risk zoning.
• Planning reservoirs, dams, canals, and hydropower sites.
• Irrigation and water allocation planning.
• Groundwater recharge zone identification and monitoring.
• Watershed management and soil erosion control.
• Water quality and pollution source mapping.
• Drought assessment.

How GIS enhances effectiveness

• Integrates many data types (rainfall, DEM, land cover, satellite images) in one platform.
• Automates processing of large areas, saving time and cost.
• Allows scenario testing, such as the effect of land-use change or climate scenarios.
• Provides clear visual outputs for decision makers and communities.
• Links with real-time sensors and remote sensing for early warning, and supports updating the model quickly.

This improves flood mitigation, water allocation, and disaster preparedness.

Opportunities

• Integrated urban management: GIS combines data on traffic, utilities, waste, housing, and services in one platform and digital twins.
• IoT and real-time data: Sensors, CCTV, and mobile data feed live maps for traffic, air quality, and energy.
• Better planning: Land-use planning, site selection, and growth modelling.
• Disaster resilience: Hazard mapping and emergency response.
• Citizen engagement: Web and mobile GIS for complaints, tax records, and participatory planning.
• AI and 3D: Automated feature extraction, 3D city models, and predictive analytics.
• Efficient services: Optimised garbage collection, public transport, and utilities.

Challenges

• Data quality and integration: Different formats, outdated data, and lack of standards between departments.
• Cost and skills: Lack of trained staff, funding, and infrastructure, especially in developing countries.
• Privacy and security: Tracking of people and cyber-attack risks.
• Interoperability: Combining legacy systems and open standards.
• Institutional issues: Poor coordination between agencies and weak data-sharing policies.
• Scalability: Handling big, real-time data volumes.
• Sustainability: Maintaining and updating the system after the initial project.

With sound policies, open standards, and capacity building, GIS can be the spatial backbone of smart cities.

Comparison

Best for global mapping: A cylindrical projection, since it shows the whole world on a rectangular map with a simple grid. Mercator suits navigation because it is conformal, but it inflates high-latitude areas, so for general world maps a compromise or equal-area cylindrical variant (Robinson, Mollweide, Gall-Peters) is preferred. Conic and azimuthal projections cannot display the entire globe without severe distortion.

Map: A map is a scaled, symbolised, and generalised graphic representation of all or part of the Earth's surface (or another space) on a flat medium, showing selected features and their spatial relationships.

Essential characteristics

• Scale: Reduction relationship between map and ground.
• Projection and coordinate system: A defined way of placing the curved surface on a plane, giving locational reference.
• Symbolisation: Features are shown with symbols, colours, and labels explained in a legend.
• Generalisation and selection: Only relevant features are shown and simplified for clarity.
• Orientation: Direction indicated by a north arrow or grid.
• Accuracy and purpose: Content suits a purpose (topographic, thematic).
• Map elements: Title, legend, scale, and source.

Difference from other spatial representations

• Unlike a photograph or satellite image, which records everything visible without selection, a map is interpreted, selective, and symbolised.
• Unlike a sketch, a map has scale, projection, and measurable accuracy.
• Unlike a globe, a map is flat, portable, and can show large scale detail, but it carries distortion.
• Unlike a digital GIS database, a map is a fixed, static presentation, whereas a GIS is dynamic and analysable.

GPS

• Advantages: Fast and direct coordinate capture, accurate (metre to centimetre with RTK), works in all weather and gives real-time positions, and is cheap with mobile devices.
• Limitations: Needs open sky, so poor under forest, in valleys, or indoors; accuracy affected by multipath and atmospheric errors; collects only point/track data and is slow for large areas.

Remote sensing

• Advantages: Covers large and inaccessible areas, repeated observation for change detection, digital format ready for GIS, multispectral information, and low cost per area.
• Limitations: Cloud cover, limited resolution for some tasks, needs processing expertise and ground verification, and high-resolution images are costly.

Surveying (total station, levelling)

• Advantages: Very high accuracy, suitable for cadastre and engineering, and gives precise detail and legal-grade data.
• Limitations: Slow, labour-intensive, and expensive for large areas; needs line of sight and trained personnel; limited coverage.

The best practice is to combine them, using remote sensing for broad coverage, GPS for field checks, and surveying for precise boundaries.

Network analysis studies flow and movement along linear networks such as roads, rivers, and pipelines, which are modelled as edges (links) joined at nodes (junctions) with attributes like length, travel time, speed, and one-way restrictions.

Process

1. Build the network dataset: Prepare clean line data with connectivity (topology) and attributes (cost, impedance, turns, barriers).
2. Define the problem and impedance: Choose the cost to minimise, such as distance or time.
3. Set inputs: Origins, destinations, stops, barriers, and constraints such as vehicle capacity or time windows.
4. Run the solver: Algorithms such as Dijkstra's compute the result. Common types are shortest/fastest path, closest facility, service area, origin-destination cost matrix, location-allocation, and vehicle routing.
5. Interpret and display the results: Routes, directions, and costs shown on a map.

Example: transportation logistics

A distribution company has a warehouse in Kathmandu and 20 delivery points. A road network with travel times, one-way streets, and traffic restrictions is used. A vehicle routing solver assigns deliveries to vehicles and finds the sequence with minimum total time, respecting capacity and delivery windows. The result is lower fuel cost and faster delivery. Closest-facility analysis can also dispatch the nearest ambulance or fire truck.

GIS database design is the process of planning how spatial data (geometry) and attribute data (descriptive information) will be organised, structured and stored in a geodatabase so that it can be captured, edited, queried and analysed efficiently. It normally passes through three levels: the conceptual design (what real-world entities, such as roads, parcels or rivers, and their relationships are needed), the logical design (feature classes, tables, geometry types, attributes, keys, relationships and topology rules) and the physical design (file or DBMS implementation, indexing, coordinate system, storage and access control).

Digital Elevation Models (DEMs) in terrain analysis

A DEM is a raster (or TIN) representation of the bare-earth surface in which each cell stores an elevation value $z$. Because it is a continuous surface, many terrain properties can be derived from it:

• Slope is the maximum rate of change of elevation between a cell and its neighbours, given by $\text{slope}=\arctan\sqrt{(\partial z/\partial x)^2+(\partial z/\partial y)^2}$. It is used for landslide susceptibility, road alignment and soil-erosion studies.
• Aspect is the compass direction the slope faces. It is used for solar exposure, agriculture and snow-melt studies.
• Hillshade and relief give a shaded 3D-looking view for cartographic display.
• Contours and profiles are generated from the DEM, together with cross-sections and line-of-sight (viewshed) analysis for towers and cell-phone coverage.
• Hydrological analysis covers flow direction, flow accumulation, stream networks and watershed delineation.
• Volume and cut-and-fill calculations are used for dams, mining and construction.
• Curvature classifies terrain into convex and concave forms.

Thus the DEM is the basic input of terrain and hydrological modelling in GIS.

(a) Components of GIS

A GIS has five main components:

1. Hardware: computers, servers, GPS receivers, scanners, digitizers, plotters and mobile devices on which the GIS runs.
2. Software: GIS packages (ArcGIS, QGIS, GRASS) providing tools for input, storage, editing, analysis and display of spatial data.
3. Data: spatial data (vector and raster) and attribute data, which is the most expensive and important component.
4. People: users, analysts, database administrators and managers who design, operate and interpret the system.
5. Methods (procedures): the rules, workflows and analysis techniques that the organisation follows to obtain correct and consistent results.

(b) TIN data structures

A Triangulated Irregular Network (TIN) is a vector-based model of a surface made of non-overlapping triangles built from irregularly spaced sample points (nodes) carrying $x,y,z$ values. Triangles are normally formed by Delaunay triangulation, which maximises the minimum angle of the triangles and avoids thin slivers.

It is stored using node, edge and triangle tables, and each triangle records its three nodes and its neighbouring triangles (topology). Each triangle is a plane facet, so slope and aspect are constant within it.

Advantages: points are dense in rugged areas and sparse in flat areas, so the storage is efficient, and breaklines such as ridges and streams can be honoured. Disadvantages: it is more complex to build and process than a raster DEM.

(c) Recent trends in GIS

• Web GIS and cloud GIS (ArcGIS Online, Google Earth Engine) and mobile GIS for field data collection.
• Integration of GIS with AI/machine learning (GeoAI) and big-data analytics.
• Use of UAV/drone mapping, LiDAR and high-resolution satellite imagery.
• Real-time and IoT-based GIS, 3D GIS, digital twins and smart cities.
• Open-source GIS and open data (OpenStreetMap), crowdsourcing and volunteered geographic information.

GIS data modeling is the process of representing real-world geographic phenomena as a structured set of data objects, relationships and rules that can be stored in and processed by a GIS. It proceeds in stages from abstract to concrete.

Process of GIS data modeling

1. Define the requirements: identify the purpose of the project, the users, the questions to be answered and the study area and scale.
2. Conceptual model: describe the real-world entities (rivers, roads, wards, buildings), their attributes and their relationships (for example, "a ward contains many buildings") in a high-level, software-independent form, often with an entity-relationship diagram.
3. Logical model: decide how the entities will be represented. Discrete objects are modelled as vector features (point, line, polygon) and continuous fields as raster grids or TINs. Define feature classes, attribute tables, data types, domains, primary and foreign keys, and topological rules (such as "polygons must not overlap").
4. Physical model: implement the logical design in a particular software and format (shapefile, file or enterprise geodatabase, PostGIS), including the coordinate reference system, indexing, storage and access rights.
5. Data population and validation: load data, check its accuracy against the model, correct errors, and document it with metadata.
6. Review and refinement: test the model with typical queries and revise it as requirements change.

Importance for spatial analysis

• It ensures that data is organised consistently, so overlay, buffer, network and proximity analyses give correct results.
• Topological rules and integrity constraints remove errors such as gaps, overlaps and dangles.
• It reduces redundancy and speeds up queries and storage.
• A good model lets different datasets be integrated and shared between users and departments.
• It keeps the system flexible and maintainable, so new data and analyses can be added later.

A poorly designed model gives wrong or unusable analysis results, however powerful the software is.

Raster-based analysis works on data stored as a regular grid of cells (pixels), where each cell holds a value representing a continuous or categorical phenomenon such as elevation, temperature, rainfall or land cover. Analysis is carried out cell by cell using map algebra.

Main types of raster analysis

• Local operations: the output value of a cell depends only on the same cell in one or more input layers (for example, reclassification, or $\text{NDVI}=(NIR-Red)/(NIR+Red)$, or a weighted overlay of suitability layers).
• Focal (neighbourhood) operations: the output depends on the cell and its neighbours in a moving window (smoothing, slope, density, filtering).
• Zonal operations: statistics are computed for each zone defined by another layer (mean rainfall per district).
• Global operations: the output depends on the entire raster (distance surfaces, cost-distance, flow accumulation, viewshed).
• Surface analysis: slope, aspect, hillshade, contouring and interpolation from a DEM.

Difference from vector-based analysis

In practice the two are combined: for example, raster suitability analysis followed by conversion of the result to vector polygons for mapping.

Surface modeling is the representation and analysis of continuous phenomena, such as terrain elevation, rainfall, temperature or pollutant concentration, as a surface in which every location has a value $z=f(x,y)$. As values are measured only at sample points, a continuous surface is created by interpolation.

Representation of surfaces

• Raster grid (DEM): a regular lattice of elevation values.
• TIN: irregular triangles from sample points and breaklines.
• Contour lines and isolines, which join places of equal value.

Interpolation methods

• Inverse Distance Weighting (IDW): the estimate is a weighted average of nearby points, $\hat z=\dfrac{\sum w_i z_i}{\sum w_i}$ with $w_i=1/d_i^{\,p}$.
• Spline: fits a smooth minimum-curvature surface through the points.
• Kriging: a geostatistical method based on a semivariogram, which also gives error estimates.
• Trend surface and natural neighbour methods.

Surface analysis operations: slope, aspect, hillshade, curvature, contouring, profiles, viewshed, volume (cut and fill) and watershed delineation.

Role in environmental analysis

• Hydrology and flood modelling: flow direction, accumulation, drainage networks and flood-prone area mapping from a DEM.
• Landslide and erosion risk: slope, aspect and curvature layers combined with rainfall and soil data.
• Climate and pollution studies: interpolated surfaces of rainfall, temperature, air-quality and groundwater levels show spatial patterns from a few stations.
• Watershed and soil conservation planning: delineating catchments and estimating runoff and sediment.
• Land suitability and habitat modelling: elevation, slope and aspect determine vegetation, wildlife and agricultural suitability.
• Visibility and impact assessment: viewshed for siting towers and for the visual impact of projects.

Surface modeling therefore turns scattered point measurements into continuous information for decisions on natural resources, hazards and environmental planning.