Topography is one of the most important factors in hydrological activities, water flow direction, sediment and contaminant transport, and irrigation process. Consequently, terrain affects crop yield, soil and water quality, and field mechanization processes. Digital representation of topography and its use in precision agriculture has been enabled by improvements in sensing and computing technologies (Westphalen et al., 2004). A Digital Elevation Model (DEM) is one of the simplest and most commonly used digital representations of the topography. In a DEM, the earth's surface is represented by spatially referenced regular grid points where each grid cell represents a ground elevation. In agriculture, DEMs play a major role in watershed modelling and hydrological flow (Renschler et al., 2002), evaluating erosion and environmental impact (Martinez-Casasnovas, 2003), and understanding spatial yield variability (Kravchenko and Bullock, 2000). During the past decade, there has been a significant increase in the production of DEMs using airborne LIDAR (light detection and ranging) (Jensen, 2000).
Lidar technology is an active remote sensing technique providing direct range measurements between the laser scanner and the Earth's surface. Such distance measurements are mapped into 3D point clouds with sub-meter vertical accuracy. This technology has emerged as a promising method for acquiring digital elevation data effectively and accurately. Since the technology is fully automated for generating digital elevation data, many researchers have paid attention to the technology and its applications (Ackermann, 1999). This situation also forces scientists or applied users who want to incorporate a DEM into their study to carefully consider the source of the DEM.
Lidar data has been applied in a multiple of disciplines, including geology, archaeology, geomorphology, engineering, resource management and disaster assessment and planning ( Bater and Coops, 2009). Recently, lidar has been gaining recognition in forestry activities such as stand characterization, forestry inventories and forest operations (Akay et. Al., 2009).
Problem Statement
LiDAR data have become a major source of digital terrain information (Raber et al., 2007) and has been used in a wide of areas, with terrain modeling being the primary focus of most LiDAR collection missions (Hodgson et al.,2005). The use of LiDAR for terrain data collection and DEM generation is the most effective way and is becoming a standard practice in spatial science community (Hodgson and Bresnahan, 2004). Although LiDAR data has become more affordable for users due to the gradually dropping of the costs of LiDAR data collection, how to effectively process the raw LiDAR data and extract useful information remains a big challenge. Furthermore, because of the specific characteristics of LiDAR data, issues such as the choices of modeling methods, interpolation algorithm, grid size, and data reduction are challenging study topics for DEM generation and quality control (Liu, 2008).
For many applications related to DEM, more accurate terrain modeling is needed to meet the requirement for terrain description. Although DEM generation from airborne LiDAR has been documented by several researchers (Lloyd and Atkinson, 2002;Wack and Wimmer, 2002;Lee,2004; Gonzales-Seco et al., 2006; Loyd and Atkinson, 2006), how to generating a high quality DEM using LiDAR data, especially in a large area is still an active research area.
Numerous studies have demonstrated that the accuracy of DEMs varies with changes in terrain and land cover type including Hodgson and Bresnahan (2004), Hodgson et. al (2005), Su and Bork (2006) and Rabel et al (2002) . As a result from lidar data collection, Hodgson and Bresnahan (2004) decomposed lidar error into three components; lidar system, horizontal interpolation and surveyor errors.
Pfeifer and Stadler (2001) assessed the derivation of DEM at the Stutgart University, Russia. Validation data consisted of a DEM generated from measurement ground points. All together four (4) area (grassland, sport ground Vaihingen, sport ground Illingen and Railway station) with different terrain characteristics have been measured. Pfeifer and Stadler (2001) reported RMS errors of 0.08 - 0.48m for different land cover class. The errors for the railway ramp are higher. This is a consequence of the structure of the elevation model that they used.
Hodgson et al.(2005) examined the effects of land cover and slope on DEM accuracy for a watershed in piedmont of North Carolina, USA. Land cover classes included grass and scrub/shrub, and pine deciduous, and mixed forests. Lidar data were collected in leaf-off conditions with an average ground return posting distance of one point every 31.1m2 (corresponding to density of 0.03 points/m). Slope was then modeled by linear interpolation of a triangulated irregular network (TIN). Reference data consisted of 1225 survey-grade points collected along 23 transects, and reference slope was calculated as the average slope of adjacent segments along survey transects. Hodgson et al. (2005) reported RMS errors of 0.145-0.361m for the different land cover classes, with higher errors occurring in areas with tall canopy vegetation. The scrub/shrub class, however, exhibited the largest RMS error. Little evidence was found for increased elevation errors in areas with slopes from 0 degree to 10 degree , but lidar-derived slope was generally under-predicted as terrain slope increased.
Currently, no significant studies have been conducted for evaluate an accuracy assessment of Lidar- Derived DEM for different land cover in Malaysia.
The land cover map for Malaysia mainly covered by forest, cropland and waterbodies. Percentage area for each land cover is 47.98 % (forest), 31.43% (cropland), 0.34% (waterbodies) and 20.25% (other). Forested and cropland area are believed to be the controlling factors to the accuracy of lidar data.
Introduce some research scenario-such as different study area, different landcover types and also different soil characteristics. These are believed to be the controlling factor, especially the landcover - structure and density
One factor that affects Lidar DEM is land cover or vegetation type. In forest areas, there are many factors and their combinations that have effect on Lidar DEM. These
factors include height of trees, forest biomass or stem volume,type of trees (coniferous, deciduous or even single tree species).
Based upon the result of this previous research, it is apparent that land
Hodgson et al. (2003) found that land cover-types were a significant factor when extracting elevation information from leaf-on lidar data in North Carolina.
But the study on Lidar DEM elevation errors for different land cover in Malaysia condition has not yet been examined
Land cover classification is a fundamental parameter describing the Earth's surface. With sufficient calibration, a land cover map can be used to identify spatial patterns of physical quantities such as carbon storage or vegetation cover as well as more abstract phenomena such
as land use. The land cover map for Malaysia mainly covered by forest, cropland and waterbodies. Percentage area for each land cover is 47.98 % (forest), 31.43% (cropland), 0.34% (waterbodies) and 20.25% (other).
Different land cover produces different errors. Previous research has shown that the accuracy of DEMs varies with changes in terrain and land cover type.
To get accurate DEM,
Research Objectives
The research objectives are as follows;
To investigate and review existing practice in generation of DEM using LIDAR for different land cover. This involved with existing application, the accuracy, processing, DEM generation methods and ……
State
application
Landcover
Accuracy
Lidar approach
To quantify for Malaysian condition
Modified existing approach
New approach
(This is a main contribution)
To compare lidar and aerial photograph of DEM generation
Accuracy
Availability
Economy
(Testing and Validation finding with aerial photograph)
Dsd
Research methodology
To meet the research objectives, the proposed research methodology is:
Review the basic theory of airborne lidar, the types of airborne lidar system available, lidar data processing, land cover and land cover classification.
Investigate and review existing practice in generation of DEM using lidar for different land cover. The investigating involved with the existing application which focusing on method of data processing, DEM generation methods, and accuracy.
Acquire airborne Lidar data and aerial photograph
Ground survey - GPS ground control point
Lidar data and aerial photograph processing
Objective no 1;
Lidar Data collection
Justification for research
In Malaysia, there have been no systematic studies to investigate the accuracy of DEM generating from Airborne Lidar Data.
Theoretically, the importance of this study are to bridge the
The results from this study would greatly benefits Department of Irrigation and Drainage Malaysia (DID), Department of Environmental Malaysia (DOE),
Example of airborne lidar land cover (related) studies
Reference Study
Location
Application/topic
Lidar -Data collection / reference data
Landcover
Categories
Accuracy
(height)/ RMSE (m)
DEM generation method
Software
Hodgson, M.E., J.R. Jensen, L. Schmidt, S. Schill, and B. Davis,
(2003).
North Carolina (USA)
An Evaluation of LIDAR- and IFSAR-derived Digital
Elevation Models in Leaf-on Conditions with USGS Level 1 and
Level 2 DEMs
-During leaf on condition (June,2000)
-flight altitude (2400m)
-Swath width (1.8km)
-Footprint size (79cm)
Reference data-
-survey reference point (1470)(GPS +conventional surveying technique
Low grass
High grass
Scrub/shrub
Pine
Deciduous
0.33
0.37
1.53
0.46
1.22
Hodgson, M.E., and P. Bresnahan, (2004).
South Carolina(USA)
-During leaf on condition
-2000 km square
-Flight altitude (1207 m)
-1 billion points (250 million points - ground)
Pavement,
Low Grass
High grass, Brush/low trees, evergreen Deciduous
0.18
0.22
0.18
0.23
0.17
0.26
TIN
Hodgson et al.(2005)
North Carolina, USA
-During leaf-Off Condition
Rabel et al. (2002)
Estern North Carolina (USA)
Vegetation Classification
-Collected by EarthData
-Using AeroScan instruments (15,000 pulse /second)
-During leaf-on condition
-3.25 km x 3.5 km
-flown at 5000 ft (1524m)
-
Pine
Deciduous
Mixed (pine or deciduous)
Scrub (<6m)
High Grass (1-2 m)
Low Grass (<1m)
Mean absolute Vertical error
0.46
2.43
2.42
1.51
0.27
0.10
TIN
Adams and Chandler (2002)
Dorset (UK)
Soft-Cliff monitoring
-Using Optech ALTM 1020
-flying height- 1000m
grassland
0.26 m
REAL software package
Pfeifer, N. and Stadler, P., 2001
Vaihingen, Stutgart (Russia)
Classification
-Optech laser scanner
Grassland
Sport ground
Railway station
0.11
0.08
0.48
SCOP ++
Su and Bork (2006)
Western Canada
DEM accuracy
Bater, C.W. and Coops, N.C (2009)
Vancouver Island, British columbia
DEM interpolation
Collected by Terra Remote Sensing
Using Mark II discrete return sensor
-ground return spacing - 1.5m square (0.7 return/meter square
Mean Obsolute errors
TIN
(slope map derived from TIN)
Kajian penyelidikan sebelum ini menunjukkan bahawa ketepatan DEM adalah berbeza mengikut perubahan terhadap topografi (terrain) dan jenis-jenis litupan (cover type) (eg. Pfeifer & stadler, 2001, Adam dan Chandler, 2002
Pfiefer dan Stadler (2001) telah menetukan ketepatan terhadap Lidar derived DEM di Vaihingin, Stutgart Russia.
Example of airborne lidar DEM studies
Reference Study
Location
application
Lidar -Data collection
Landcover
Categories
Accuracy
(height)/ RMSE (m)
DEM generation method
Software
Hodgson, M.E., and P. Bresnahan, (2004).
South Carolina(USA)
-During leaf on condition
-2000 km square
-Flight altitude (1207 m)
-1 billion points (250 million points - ground)
Pavement,
Low Grass
High grass, Brush/low trees, evergreen Deciduous
0.18
0.22
0.18
0.23
0.17
0.26
TIN
Chapter 2: Background theory
2.1 Lidar technology
2.1.1 Basic theory of Irborne Lidar
2.1.2 Airborne Lidar systems
2.2 Landcover Categories
2.3 DEM
Hodgson, M.E., and P. Bresnahan, 2004. Accuracy of airborne LIDAR-derived elevation:
Empirical assessment and error budget, Photogrammetric Engineering & Remote
Sensing, 70(3), 331-339.
Research Plan and Schedule
YEAR
2009
2010
2011
2012
MONTH
M
A
M
J
J
A
S
O
N
J
F
M
A
M
J
J
A
S
O
N
D
J
F
M
A
M
J
J
A
S
A
N
D
J
F
M
Literature Review &Proposal
12 month
Investigate and review existing practice in generation of DEM
5 months
Acquiring LiDAR data and aerial photograph
5 months
Field Work- GPS GCP (ground control point)
6 months
10 months
Thesis Writing
24 months
