Today’s post continues where “Why you should be using PostGIS trajectories” leaves off. It’s the result of a collaboration with Eva Westermeier. I had the pleasure to supervise her internship at AIT last year and also co-supervised her Master’s thesis [0] on the topic of enriching trajectories with information about their geographic context.

Context-aware analysis of movement data is crucial for different domains and applications, from transport to ecology. While there is a wealth of data, efficient and user-friendly contextual trajectory analysis is still hampered by a lack of appropriate conceptual approaches and practical methods. (Westermeier, 2018)

Part of the work was focused on evaluating different approaches to adding context information from vector datasets to trajectories in PostGIS. For example, adding land cover context to animal movement data or adding information on anchoring and harbor areas to vessel movement data.

Classic point-based model vs. line-based model

The obvious approach is to intersect the trajectory points with context data. This is the classic point data model of contextual trajectories. It’s straightforward to add context information in the point-based model but it also generates large numbers of repeating annotations. In contrast, the line data model using, for example, PostGIS trajectories (LinestringM) is more compact since trajectories can be split into segments at context borders. This creates one annotation per segment and the individual segments are convenient to analyze (as described in part #12).

Spatio-temporal interpolation as provided by the line data model offers additional advantages for the analysis of annotated segments. Contextual segments start and end at the intersection of the trajectory linestring with context polygon borders. This means that there are no gaps like in the point-based model. Consequently, while the point-based model systematically underestimates segment length and duration, the line-based approach offers more meaningful segment length and duration measurements.

Schematic illustration of a subset of an annotated trajectory in two context classes, a) systematic underestimation of length or duration in the point data model, b) full length or duration between context polygon borders in the line data model (source: Westermeier (2018))

Another issue of the point data model is that brief context changes may be missed or represented by just one point location. This makes it impossible to compute the length or duration of the respective context segment. (Of course, depending on the application, it can be desirable to ignore brief context changes and make the annotation process robust towards irrelevant changes.)

Schematic illustration of context annotation for brief context changes, a) and b)
two variants for the point data model, c) gapless annotation in the line data model (source: Westermeier (2018) based on Buchin et al. (2014))

Beyond annotations, context can also be considered directly in an analysis, for example, when computing distances between trajectories and contextual point objects. In this case, the point-based approach systematically overestimates the distances.

Schematic illustration of distance measurement from a trajectory to an external
object, a) point data model, b) line data model (source: Westermeier (2018))

The above examples show that there are some good reasons to dump the classic point-based model. However, the line-based model is not without its own issues.


Computing the context annotations for trajectory segments is tricky. The main issue is that ST_Intersection drops the M values. This effectively destroys our trajectories! There are ways to deal with this issue – and the corresponding SQL queries are published in the thesis (p. 38-40) – but it’s a real bummer. Basically, ST_Intersection only provides geometric output. Therefore, we need to reconstruct the temporal information in order to create usable trajectory segments.

Finally, while the line-based model is well suited to add context from other vector data, it is less useful for context data from continuous rasters but that was beyond the scope of this work.


After the promising results of my initial investigations into PostGIS trajectories, I was optimistic that context annotations would be a straightforward add-on. The line-based approach has multiple advantages when it comes to analyzing contextual segments. Unfortunately, generating these contextual segments is much less convenient and also slower than I had hoped. Originally, I had planned to turn this work into a plugin for the Processing toolbox but the results of this work motivated me to look into other solutions. You’ve already seen some of the outcomes in part #20 “Trajectools v1 released!”.


[0] Westermeier, E.M. (2018). Contextual Trajectory Modeling and Analysis. Master Thesis, Interfaculty Department of Geoinformatics, University of Salzburg.

This post is part of a series. Read more about movement data in GIS.


If you’ve been following my posts, you’ll no doubt have seen quite a few flow maps on this blog. This tutorial brings together many different elements to show you exactly how to create a flow map from scratch. It’s the result of a collaboration with Hans-Jörg Stark from Switzerland who collected the data.

The flow data

The data presented in this post stems from a survey conducted among public transport users, especially commuters (available online at: Among other questions, the questionnair asks where the commuters start their journey and where they are heading.

The answers had to be cleaned up to correct for different spellings, spelling errors, and multiple locations in one field. This cleaning and the following geocoding step were implemented in Python. Afterwards, the flow information was aggregated to count the number of nominations of each connection between different places. Finally, these connections (edges that contain start id, destination id and number of nominations) were stored in a text file. In addition, the locations were stored in a second text file containing id, location name, and co-ordinates.

Why was this data collected?

Besides travel demand, Hans-Jörg’s survey also asks participants about their coffee consumption during train rides. Here’s how he tells the story behind the data:

As a nearly daily commuter I like to enjoy a hot coffee on my train rides. But what has bugged me for a long time is the fact the coffee or hot beverages in general are almost always served in a non-reusable, “one-use-only-and-then-throw-away” cup. So I ended up buying one of these mostly ugly and space-consuming reusable cups. Neither system seem to satisfy me as customer: the paper-cup produces a lot of waste, though it is convenient because I carry it only when I need it. With the re-usable cup I carry it all day even though most of the time it is empty and it is clumsy and consumes the limited space in bag.

So I have been looking for a system that gets rid of the disadvantages or rather provides the advantages of both approaches and I came up with the following idea: Installing a system that provides a re-usable cup that I only have with me when I need it.

In order to evaluate the potential for such a system – which would not only imply a material change of the cups in terms of hardware but also introduce some software solution with the convenience of getting back the necessary deposit that I pay as a customer and some software-solution in the back-end that handles all the cleaning, distribution to the different coffee-shops and managing a balanced stocking in the stations – I conducted a survey

The next step was the geographic visualization of the flow data and this is where QGIS comes into play.

The flow map

Survey data like the one described above is a common input for flow maps. There’s usually a point layer (here: “nodes”) that provides geographic information and a non-spatial layer (here: “edges”) that contains the information about the strength or weight of a flow between two specific nodes:

The first step therefore is to create the flow line features from the nodes and edges layers. To achieve our goal, we need to join both layers. Sounds like a job for SQL!

More specifically, this is a job for Virtual Layers: Layer | Add Layer | Add/Edit Virtual Layer

SELECT StartID, DestID, Weight, 
       make_line(a.geometry, b.geometry)
FROM edges
JOIN nodes a ON edges.StartID = a.ID
JOIN nodes b ON edges.DestID = b.ID
WHERE a.ID != b.ID 

This SQL query joins the geographic information from the nodes table to the flow weights in the edges table based on the node IDs. In the last line, there is a check that start and end node ID should be different in order to avoid zero-length lines.

By styling the resulting flow lines using data-driven line width and adding in some feature blending, it’s possible to create some half decent maps:

However, we can definitely do better. Let’s throw in some curved arrows!

The arrow symbol layer type automatically creates curved arrows if the underlying line feature has three nodes that are not aligned on a straight line.

Therefore, to turn our straight lines into curved arrows, we need to add a third point to the line feature and it has to have an offset. This can be achieved using a geometry generator and the offset_curve() function:


Additionally, to achieve the effect described in New style: flow map arrows, we extend the geometry generator to crop the lines at the beginning and end:

      buffer(start_point($geometry), 0.01)
   buffer(end_point( $geometry), 0.01)

By applying data-driven arrow and arrow head sizes, we can transform the plain flow map above into a much more appealing map:

The two different arrow colors are another way to emphasize flow direction. In this case, orange arrows mark flows to the west, while blue flows point east.

 x(start_point($geometry)) - x(end_point($geometry)) < 0


As you can see, virtual layers and geometry generators are a powerful combination. If you encounter performance problems with the virtual layer, it’s always possible to make it permanent by exporting it to a file. This will speed up any further visualization or analysis steps.

This post looks into the current AI hype and how it relates to geoinformatics in general and movement data analysis in GIS in particular. This is not an exhaustive review but aims to highlight some of the development within these fields. There are a lot of references in this post, including some to previous work of mine, so you can dive deeper into this topic on your own.

I’m looking forward to reading your take on this topic in the comments!

Introduction to AI

The dream of artificial intelligence (AI) that can think like a human (or even outsmart one) reaches back to the 1950s (Fig. 1, Tandon 2016). Machine learning aims to enable AI. However, classic machine learning approaches that have been developed over the last decades (such as: decision trees, inductive logic programming, clustering, reinforcement learning, neural networks, and Bayesian networks) have failed to achieve the goal of a general AI that would rival humans. Indeed, even narrow AI (technology that can only perform specific tasks) was mostly out of reach (Copeland 2018).

However, recent increases in computing power (be it GPUs, TPUs or CPUs) and algorithmic advances, particularly those based on neural networks, have made this dream (or nightmare) come closer (Rao 2017) and are fueling the current AI hype. It should be noted that artificial neural networks (ANN) are not a new technology. In fact, they used to be not very popular because they require large amounts of input data and computational power. However, in 2012, Andrew Ng at Google managed to create large enough neural networks and train them with massive amounts of data, an approach now know as deep learning (Copeland 2018).

Fig. 1: The evolution of artificial intelligence, machine learning, and deep learning. (Image source: Tandon 2016)

Machine learning & GIS

GIScience or geoinformatics is not new to machine learning. The most well-known application is probably supervised image classification, as implemented in countless commercial and open tools. This approach requires labeled training and test data (Fig. 2) to learn a prediction model that can, for example, classify land cover in remote sensing imagery. Many classification algorithms have been introduced, ranging from maximum likelihood classification to clustering (Congedo 2016) and neural networks.

Fig. 2: With supervised machine learning, the algorithm learns from labeled data. (Image source: Salian 2018)

Like in other fields, neural networks have intrigued geographers and GIScientists for a long time. For example, Hewitson & Crane (1994) state that “Neural nets offer a fascinating new strategy for spatial analysis, and their application holds enormous potential for the geographic sciences.” Early uses of neural network in GIScience include, for example: spatial interaction modeling (Openshaw 1998) and hydrological modeling of rainfall runoff (Dawson & Wilby 2001). More recently, neural networks and deep learning have enabled object recognition in georeferenced images. Most prominently, the research team at Mapillary (2016-2019) works on object recognition in street-level imagery (including fusion with other spatial data sources). Even Generative adversarial networks (GANs) (Fig. 3) have found their application in GIScience: for example, Zhu et al. (2017) (at the Berkeley AI Research (BAIR) laboratory) demonstrate how GANs can generate road maps from aerial images and vice versa, and Zhu et al. (2019) generate artificial digital elevation models.

Fig. 3: In a GAN, the discriminator is shown images from both the generator and from the training dataset. The discriminator is tasked with determining which images are real, and which are fakes from the generator. (Image source: Salian 2018)

However, besides general excitement about new machine learning approaches, researchers working on spatial analysis (Openshaw & Turton 1996) caution that “conventional classifiers, as provided in statistical packages, completely ignore most of the challenges of spatial data classification and handle a few inappropriately from a geographical perspective”. For example, data transformation using principal component or factor scores is sensitive to non-normal data distribution common in geographic data and many methods ignore spatial autocorrelation completely (Openshaw & Turton 1996). And neural networks are no exception: Convolutional neural networks (CNNs) are generally regarded appropriate for any problem involving pixels or spatial representations. However, Liu et al. (2018) demonstrate that they fail even for the seemingly trivial coordinate transform problem, which requires learning a mapping between coordinates in (x, y) Cartesian space and coordinates in one-hot pixel space.

The integration of spatial data challenges into machine learning is an ongoing area of research, for example in geostatistics (Hengl & Heuvelink 2019).

Machine learning and movement data

More and more movement data of people, vehicles, goods, and animals is becoming available. Developments in intelligent transportation systems specifically have been sparked by the availability of cheap GPS receivers and many models have been built that leverage floating car data (FCD) to classify traffic situations (for example, using visual analysis (Graser et al. 2012)), predict traffic speeds (for example, using linear regression models (Graser et al. 2016)), or detect movement anomalies (for example, using Gaussian mixture models (Graser & Widhalm 2018)). Beyond transportation, Valletta et al. (2017) describe applications of machine learning in animal movement and behavior.

Of course deep learning is making its way into movement data analysis as well. For example, Wang et al. (2018) and Kudinov (2018) trained neural networks to predict travel times in a transport networks. In contrast to conventional travel time prediction models (based on street graphs with associated speeds or travel times), these are considerably more computationally intensive. Kudinov (2018) for example, used 300 million simulated trips (start and end location, start time, and trip duration) as input and “spent about eight months of running one of the GP100 cards 24-7 in a search for an efficient architecture, spatial and statistical distributions of the training set, good values for multiple hyperparameters”.  More recently, Zhang et al. (2019) (at Microsoft Research Asia) used deep learning to predict flows in spatio-temporal networks. It remains to be seen if deep learning will manage to out-perform classical machine learning approaches for predictions in the transportation sector.

What would a transportation AI look like? Would it be able to drive a car and follow data-driven route recommendations (e.g. from or would it purposefully ignore them because other – more basic systems – blindly follow it? Logistics AI might build on these kind of systems while simultaneously optimizing large fleets of vehicles. Transport planning AI might replace transport planners by providing reliable mobility demand predictions as well as resulting traffic models for varying infrastructure and policy scenarios.


The opportunities for using ML in geoinformatics are extensive and have been continuously explored for a multitude of different research problems and applications (from land use classification to travel time prediction). Geoinformatics is largely playing catch-up with the quick development in machine learning (including deep learning) that promise new and previously unseen possibilities. At the same time, it is necessary that geoinformatics researchers are aware of the particularities of spatial data, for example, by developing models that take spatial autocorrelation into account. Future research in geoinformatics should incorporate learnings from geostatistics to ensure that resulting machine learning models incorporate the geographical perspective.


  • Congedo, L. (2016). Semi-Automatic Classification Plugin Documentation. DOI:
  • Copeland, M. (2016) What’s the Difference Between Artificial Intelligence, Machine Learning, and Deep Learning?
  • Dawson, C. W., & Wilby, R. L. (2001). Hydrological modelling using artificial neural networks. Progress in physical Geography, 25(1), 80-108.
  • Graser, A., Ponweiser, W., Dragaschnig, M., Brandle, N., & Widhalm, P. (2012). Assessing traffic performance using position density of sparse FCD. In Intelligent Transportation Systems (ITSC), 2012 15th International IEEE Conference on (pp. 1001-1005). IEEE.
  • Graser, A., Leodolter, M., Koller, H., & Brändle, N. (2016) Improving vehicle speed estimates using street network centrality. International Journal of Cartography. doi:10.1080/23729333.2016.1189298.
  • Graser, A., & Widhalm, P. (2018). Modelling Massive AIS Streams with Quad Trees and Gaussian Mixtures. In: Mansourian, A., Pilesjö, P., Harrie, L., & von Lammeren, R. (Eds.), 2018. Geospatial Technologies for All : short papers, posters and poster abstracts of the 21th AGILE Conference on Geographic Information Science. Lund University 12-15 June 2018, Lund, Sweden. ISBN 978-3-319-78208-9. Accessible through
  • Hengl, T. Heuvelink, G.B.M. (2019) Workshop on Machine learning as a framework for predictive soil mapping
  • Hewitson, B., Crane, R. G. (Eds.) (1994) Neural Nets: Applications in Geography. Springer.
  • Kudinov, D. (2018) Predicting travel times with artificial neural network and historical routes.
  • Liu, R., Lehman, J., Molino, P., Such, F. P., Frank, E., Sergeev, A., & Yosinski, J. (2018). An intriguing failing of convolutional neural networks and the coordconv solution. In Advances in Neural Information Processing Systems (pp. 9605-9616).
  • Mapillary Research (2016-2019) publications listed on
  • Openshaw, S., & Turton, I. (1996). A parallel Kohonen algorithm for the classification of large spatial datasets. Computers & Geosciences, 22(9), 1019-1026.
  • Openshaw, S. (1998). Neural network, genetic, and fuzzy logic models of spatial interaction. Environment and Planning A, 30(10), 1857-1872.
  • Rao, R. C.S. (2017) New Product breakthroughs with recent advances in deep learning and future business opportunities.
  • Salian, I. (2018) SuperVize Me: What’s the Difference Between Supervised, Unsupervised, Semi-Supervised and Reinforcement Learning?
  • Tandon, K. (2016) AI & Machine Learning: The evolution, differences and connections
  • Valletta, J. J., Torney, C., Kings, M., Thornton, A., & Madden, J. (2017). Applications of machine learning in animal behaviour studies. Animal Behaviour, 124, 203-220.
  • Wang, D., Zhang, J., Cao, W., Li, J., & Zheng, Y. (2018). When will you arrive? estimating travel time based on deep neural networks. In Thirty-Second AAAI Conference on Artificial Intelligence.
  • Zhang, J., Zheng, Y., Sun, J., & Qi, D. (2019). Flow Prediction in Spatio-Temporal Networks Based on Multitask Deep Learning. IEEE Transactions on Knowledge and Data Engineering.
  • Zhu, J. Y., Park, T., Isola, P., & Efros, A. A. (2017). Unpaired image-to-image translation using cycle-consistent adversarial networks. In Proceedings of the IEEE international conference on computer vision (pp. 2223-2232).
  • Zhu, D., Cheng, X., Zhang, F., Yao, X., Gao, Y., & Liu, Y. (2019). Spatial interpolation using conditional generative adversarial neural networks. International Journal of Geographical Information Science, 1-24.

This post is part of a series. Read more about movement data in GIS.

MovingPandas is my attempt to provide a pure Python solution for trajectory data handling in GIS. MovingPandas provides trajectory classes and functions built on top of GeoPandas. 

To lower the entry barrier to getting started with MovingPandas, there’s now an interactive iPython notebook hosted on MyBinder. This notebook provides all the necessary imports and demonstrates how to create a Trajectory object.

Launch MyBinder for MovingPandas to get started!

PyQGIS scripts are great to automate spatial processing workflows. It’s easy to run these scripts inside QGIS but it can be even more convenient to run PyQGIS scripts without even having to launch QGIS. To create a so-called “stand-alone” PyQGIS script, there are a few things that need to be taken care of. The following steps show how to set up PyCharm for stand-alone PyQGIS development on Windows10 with OSGeo4W.

An essential first step is to ensure that all environment variables are set correctly. The most reliable approach is to go to C:\OSGeo4W64\bin (or wherever OSGeo4W is installed on your machine), make a copy of qgis-dev-g7.bat (or any other QGIS version that you have installed) and rename it to pycharm.bat:

Instead of launching QGIS, we want that pycharm.bat launches PyCharm. Therefore, we edit the final line in the .bat file to start pycharm64.exe:

In PyCharm itself, the main task to finish our setup is configuring the project interpreter:

First, we add a new “system interpreter” for Python 3.7 using the corresponding OSGeo4W Python installation.

To finish the interpreter config, we need to add two additional paths pointing to QGIS\python and QGIS\python\plugins:

That’s it! Now we can start developing our stand-alone PyQGIS script.

The following example shows the necessary steps, particularly:

  1. Initializing QGIS
  2. Initializing Processing
  3. Running a Processing algorithm
import sys

from qgis.core import QgsApplication, QgsProcessingFeedback
from qgis.analysis import QgsNativeAlgorithms

QgsApplication.setPrefixPath(r'C:\OSGeo4W64\apps\qgis-dev', True)
qgs = QgsApplication([], False)

# Add the path to processing so we can import it next
# Imports usually should be at the top of a script but this unconventional 
# order is necessary here because QGIS has to be initialized first
import processing
from processing.core.Processing import Processing

feedback = QgsProcessingFeedback()

rivers = r'D:\Documents\Geodata\NaturalEarthData\Natural_Earth_quick_start\10m_physical\ne_10m_rivers_lake_centerlines.shp'
output = r'D:\Documents\Geodata\temp\danube3.shp'
expression = "name LIKE '%Danube%'"

danube =
    {'INPUT': rivers, 'EXPRESSION': expression, 'OUTPUT': output},


When QGIS 3.0 was release, I published a Processing script template for QGIS3. While the script template is nicely pythonic, it’s also pretty long and daunting for non-programmers. This fact didn’t go unnoticed and Nathan Woodrow in particular started to work on a QGIS enhancement proposal to improve the situation and make writing Processing scripts easier, while – at the same time – keeping in line with common Python styles.

While the previous template had 57 lines of code, the new template only has 26 lines – 50% less code, same functionality! (Actually, this template provides more functionality since it also tracks progress and ensures that the algorithm can be cancelled.)

from qgis.processing import alg
from qgis.core import QgsFeature, QgsFeatureSink

@alg(name="split_lines_new_style","Alg name"), group="examplescripts","Example Scripts"))
@alg.input(type=alg.SOURCE, name="INPUT", label="Input layer")
@alg.input(type=alg.SINK, name="OUTPUT", label="Output layer")
def testalg(instance, parameters, context, feedback, inputs):
    Description goes here. (Don't delete this! Removing this comment will cause errors.)
    source = instance.parameterAsSource(parameters, "INPUT", context)

    (sink, dest_id) = instance.parameterAsSink(
        parameters, "OUTPUT", context,
        source.fields(), source.wkbType(), source.sourceCrs())

    total = 100.0 / source.featureCount() if source.featureCount() else 0
    features = source.getFeatures()
    for current, feature in enumerate(features):
        if feedback.isCanceled():
        out_feature = QgsFeature(feature)
        sink.addFeature(out_feature, QgsFeatureSink.FastInsert)
        feedback.setProgress(int(current * total))

    return {"OUTPUT": dest_id}

The key improvement are the new decorators that turn an ordinary function (such as testalg in the template) into a Processing algorithm. Decorators start with @ and are written above a function definition. The @alg decorator declares that the following function is a Processing algorithm, defines its name and assigns it to an algorithm group. The @alg.input decorator creates an input parameter for the algorithm. Similarly, there is a @alg.output decorator for output parameters.

For a longer example script, check out the original QGIS enhancement proposal thread!

For now, this new way of writing Processing scripts is only supported by QGIS 3.6 but there are plans to back-port this improvement to 3.4 once it is more mature. So give it a try and report back!

In previous posts, I already wrote about Trajectools and some of the functionality it provides to QGIS Processing including:

There are also tools to compute heading and speed which I only talked about on Twitter.

Trajectools is now available from the QGIS plugin repository.

The plugin includes sample data from MarineCadastre downloads and the Geolife project.

Under the hood, Trajectools depends on GeoPandas!

If you are on Windows, here’s how to install GeoPandas for OSGeo4W:

  1. OSGeo4W installer: install python3-pip
  2. Environment variables: add GDAL_VERSION = 2.3.2 (or whichever version your OSGeo4W installation currently includes)
  3. OSGeo4W shell: call C:\OSGeo4W64\bin\py3_env.bat
  4. OSGeo4W shell: pip3 install geopandas (this will error at fiona)
  5. From download Fiona-1.7.13-cp37-cp37m-win_amd64.whl
  6. OSGeo4W shell: pip3 install path-to-download\Fiona-1.7.13-cp37-cp37m-win_amd64.whl
  7. OSGeo4W shell: pip3 install geopandas
  8. (optionally) From download Rtree-0.8.3-cp37-cp37m-win_amd64.whl and pip3 install it

If you want to use this functionality outside of QGIS, head over to my movingpandas project!

%d bloggers like this: