**6. Discussion**

The analysis of the changes of urban form including buildings and trees in 3D space is important because contributions of changes in tree canopy cover compared with changes in grass cover have significantly different effects related to the reduction of air pollution and also evapotranspiration levels that affect the microclimate conditions. This information will result in better data driven decision making for smart solutions for future 3D developments of cities, as well as demonstrate 3D green spaces in cities if land conversion from tree cover to built-up areas occurs. Having the changes of height data, urban planners can make better decisions for smart changes of urban form for future developments, considering the negative impacts of buildings on the natural environment of the cities. Smart changes of urban form refer to data driven decision making on the areas within cities with higher potential of change.

This research introduced a robust method for exploring 3D changes of cities to see whether these changes can be considered as being toward or away from a more sustainable 3D urban form, based on the relationship between built up and vegetated areas, in particular tree canopy. Future work will focus on the speed of change in different urban areas for the vegetation and building classes.

Similar metrics can be proposed for vegetation or a combination of vegetation and buildings. However, current technical problems of data collection and also extraction of individual trees using airborne lidar need to be resolved. The data collection problem here refers to the separation of tree canopy from vegetation, as well as the loss of many lidar points in tree structures below canopy level, that affects the extraction of the complete shape of a tree crown. The incompleteness of tree shape reconstruction necessarily affects the results of the calculation of the volume of trees. In addition, the pixel-based approach in tree volume calculation may be affected by the lidar sampling density and distribution, since a tree sampled by multi-temporal airborne lidar will be represented by totally different sets of points in each epoch. As well, depending on the scattering characteristics of the lidar laser, we could see differences in the heights of pixels representing the trees in each epoch.

In the previous work, the authors found that for monitoring of 3D building changes over time, a change detection approach is preferred compared to a process of building classification of time series lidar data [29]. In this study, we found that change detection is not sufficiently accurate for vegetation classes, therefore, we recommend voxelization of the classified vegetation for future studies.

These sampling problems impact on the voxelization of the tree points. In addition, a further problem for the voxelization of lidar point clouds is that tree trunks may not be sampled leading to gaps between tree crowns and the ground which are not registered in the rasterization process.

So far, we have found that for the calculation of volumetrics of vegetation, the change detection approach is not sufficiently reliable as unchanged trees are sometimes detected as changed trees (cut trees or new trees) from the sampling of airborne lidar data. To alleviate this problem and to overcome the problem of inconsistency of classified time series lidar data, we have tested voxelization of classified buildings and vegetation. In addition, change detection of vegetation to identify growth is not sufficiently accurate. Therefore, future work should focus on development of methods for detection of vegetation growth as another class of change in urban areas. This should be possible as future lidar systems will provide much higher densities of sampling, which together with multispectral lidar systems will enable more accurate height measurement of trees.

Vegetation cover as a valuable component of a healthy built environment within cities motivates us to ask a critical question for future studies. Indeed, the question is whether 3D vegetation cover change could be a more influential factor in changing the local climate and affecting health and wellbeing than grassed areas. So far, we know that the observation of the trees in providing shade and the mechanism of photosynthesis in absorbing CO2 confirm that preserving trees within built-up

**65**

*New Metrics for Spatial and Temporal 3D Urban Form Sustainability Assessment Using Time…*

areas is crucial for caring for the environment leading toward more sustainable

Our work is intended to provide fundamental information for "wise management of natural resources" ([30], p.6) in which we provide information on the location of changes in the natural and human-made resources within cities.

From an urban planning and design perspective, this work provides a pathway to the development of rapid assessment protocols for estimating the directions in which our precincts and cities are trending in terms of sustainable development. Such an approach would take the increase (or decrease) of vegetation, in particular tree canopy, as a proxy for a rising or falling level of sustainability in an urban location. Of course, sustainability, particularly in the urban context, is far more complex than this, including such factors as climate change mitigation and adaptation, consumption of renewable and non-renewable resources, biodiversity protection and so on. But an increase or decline in urban vegetation over time represents a fundamental trend, which can indicate a great deal about other aspects of urban sustainability.

The lack of appropriate metrics and methodologies for the assessment of the sustainability of urban form over time motivated us to conduct this research. In this chapter, we proposed novel 3D metrics and potentially appropriate methodologies and tested these methods. Our tests results show that while airborne lidar data is a very accurate and promising source of data for detection of changes of the volume of buildings, there are areas requiring improvement in 3D reconstruction of trees and vegetation from airborne lidar, to enable its use in an urban change detection studies. We found integration of SVM and DSM differencing appropriate for the building change detection and proposed a voxelization approach for vegetation

We observed a trend of decreasing volume of vegetation and increasing the volume of buildings using three of epochs airborne lidar data from 2005 to 2013 in a developing urban area in Sydney, Australia. This implies a trend against sustainability goals and suggests a need for intervention policies for preserving the natural

*DOI: http://dx.doi.org/10.5772/intechopen.89617*

urban forms for future generations.

change detection in future research.

objects such as trees in a built environment.

**7. Conclusions**

*New Metrics for Spatial and Temporal 3D Urban Form Sustainability Assessment Using Time… DOI: http://dx.doi.org/10.5772/intechopen.89617*

areas is crucial for caring for the environment leading toward more sustainable urban forms for future generations.

Our work is intended to provide fundamental information for "wise management of natural resources" ([30], p.6) in which we provide information on the location of changes in the natural and human-made resources within cities.

From an urban planning and design perspective, this work provides a pathway to the development of rapid assessment protocols for estimating the directions in which our precincts and cities are trending in terms of sustainable development. Such an approach would take the increase (or decrease) of vegetation, in particular tree canopy, as a proxy for a rising or falling level of sustainability in an urban location. Of course, sustainability, particularly in the urban context, is far more complex than this, including such factors as climate change mitigation and adaptation, consumption of renewable and non-renewable resources, biodiversity protection and so on. But an increase or decline in urban vegetation over time represents a fundamental trend, which can indicate a great deal about other aspects of urban sustainability.
