In the second half of the 20th and beginning of the 21st century the amount of natural disasters has increased rapidly. Due to this rise in occurrences, more people are affected. An important indicator for people affected is the amount of damage to buildings. To gather this information aid workers now have to go into the field to gather data on the amount of destruction. In response to the possible dangers these people encounter in the field, remote sensing and analysis techniques have been developed for automated damage detection. However, due to various limitations on the implementation, these techniques are not yet widely adopted in emergency response and humanitarian aid.

This work compares two methods and two data sources for the detection of building damage. The methods are evaluated on their accuracy and implementability within humanitarian aid in disaster situations. The main methods considered are equalisation of histograms of pre-event and post-event imagery, followed by Univariate Image Differencing; and a convolutional neural network on features withdrawn from post-event imagery, using OpenStreetMap data. Remotely sensed data sources considered are synthetic aperture radar and very high resolution optical imagery. All results are analysed and compared to current standards in damage detection.

From the results it can be concluded that more research is required for a practical implementation of deep learning techniques. The constraint posed by the requirement of large datasets, make these methods impracticable without sufficient preparation and resources. More simpler methods, like Univariate Image Differencing, can be validated on smaller ground-truth datasets, and are therefore easier in implementation when resources are limited. The possible accuracy increase of deep learning methods does, at this moment, not outweigh the ease of an elementary differencing approach.

The continuously increasing population within cities imposes the future challenges related to planning and managing the sustainable environment, where people’s health and wellbeing is prioritized. Currently more than half of the world’s population live in cities which results in the rise of the human footprint, affecting the local climate. Simultaneously the planet’s climate is changing towards less predictable weather conditions with high extremes, e.g., heavy rainfalls, followed by long dry periods or severe heatwaves. Frequently extreme weather conditions are associated with flooding and hurricanes, while heatwaves represent an equally important danger for the health of city’s citizens.
UHI has been recognized as one of the leading environmental issues of the 21st century. The UHI is defined as the area within a city with the higher surface or air temperatures compared to its surroundings. The higher urban temperatures are resulting in health-related issues among the population, greater energy demands, and various economic losses. Thus, the current work is focusing on researching the air temperatures in the urban canopy layer, which are mainly affected by the heat radiated from the urban surfaces during the night. Moreover, there are differences in the heat exchange phases between the neighborhoods, which are caused by the heterogenicity of the morphological characteristics of the city. Such information can be of great use for the development of UHI mitigation strategies, but it is currently very sparse or not detailed enough. Therefore this master thesis investigates the possibility for generation of more detailed models, taking into account the intraurban variability. For the development of the statistical models explaining the UHI effect in the city of the Hague different spatial and sensor datasets have been used. The data about the temperatures in the city for 2017 have been collected with the means of the Netatmo weather stations. Additionally, different spatial representations and their effect on the statistical analysis have been investigated. In combination with the different spatial models, six distinct UHI contributing factors have been researched, namely the Buildings density, the Land cover index, the Vegetation index, the Sky View Factor, the Non-permeable surfaces in the city and the Vehicle traffic density. These variables have been calculated and utilized in the statistical analysis of their relationship with the air temperatures in the Hague. The results indicated a weak relationship between the air temperatures and the different spatial characteristics of the city where only the Sky View Factor and the Non-permeable surfaces proved to be statistically significant variables.

Amsterdam circular economy ambitions raise the question of what would be the impact of an organic waste collection system on inhabitants. Impact is an effect of a source on a receptor. Existing impact modelling studies describe impact either from source or from receptor perspective. In spatial planning, both source and receptor are variant over space. However, incorporation of the spatial dimension in existing impact modelling studies is limited, especially in presentation of outcomes.

The objective of this research is to answer: what spatio-temporal method is suitable to assess the impacts of organic waste collection system scenarios in Amsterdam? Tools used are QGIS and GRASS GIS, combined with PyQGIS scripting.

Five impact indicators were defined: three nuisance indicators (noise, odour and congestion), all quantified in three factors (intensity, temporality and affected population), and two sustainable planning indicators (CO2 emission and waste collection), quantified in intensity factor only. Calculations are performed using simple theoretical principles, implemented from both source and receptor point of view.

It was found that basic spatial tools and calculation principles suffice to quantify impact spatio-temporally. For congestion and waste collection modelling network calculations are needed. All indicator outcomes except for CO2 emission are relevant to present spatially. Both source and receptor point of view have their advantages and challenges, and therefore are complementing each other. Aggregation of factor values into one results in loss of information and transparency, which is contradicting to information and transparency gained by the high granularity spatio-temporal model. However, overview and simplicity in presentation of outcomes could be maintained by describing all impact factors per indicator in one clock-like marker. Impact assessment outcomes should be presented for individual sources and receptors. When presenting them on small scale, spatial aggregation into total values per neighbourhood is needed as well. Compared to presentation in a non-spatial graph, the spatial presentation of outcomes is of added value.

There are several 3D city models available openly, worldwide. These models are used in various applications, from which many expects a homogeneous Level of Detail (LoD). Validating the accuracy of the LoD of a model requires the inference its LoD class and its conformance to the real-world object. This process quickly becomes infeasible for large models when done manually. Yet there is no automatic method for LoD inference and validation. Therefore the thesis proposes a method to automatically infer the geometric LoD (LoD0-2.3) in 3D city models.

A central aspect of this work is the use of machine-learning to classify building models based on their LoD. It follows the assumption that a process is possible where a classifier trained in a synthetic 3D city model containing all LoD classes, and applied in real city model. Therefore ten geometry measures (features) are computed from the objects and tested with six classification algorithms. The six experiments the transferability of a classifier from the synthetic city model to the real one, multi-class (LoD0-2.3) and binary (LoD2 or not) classification, and the effect of LoD class imbalance by introducing various amounts of LoD1 objects into the LoD2 model. Furthermore, by using a point cloud as ground truth, this explored the possibility of validating the inferred LoD classes.

The results indicate that the classifier is not transferable to the real data set when trained on the synthetic city model, which is probably due to the significant difference in object shapes between the two models. Binary classification outperforms the multi-class case and it is favourable for LoD validation where the main question is whether the model conforms the stated LoD or not. Finally, class-imbalance can reduce the classification with as much as 20%.