Programme

Agenda

Tuesday, 28 October 2025

8:45-10:00	Workshop session 0: Gentle introduction in the archaeological interpretation of lidar derived data
10:00-10:30	break
10:30-11:45	Workshop session 0: Gentle introduction in the archaeological interpretation of lidar derived data
12:00-13:15	lunch
13:30-14:00	Introduction to the meeting
14:00-16:00	Keynote 1 by Dave Cowley Attrition, palaeo-landscapes, and the patterns of the past in farmed landscapes Keynote 2 by João Fonte Do we really need drone-derived lidar data in landscape archaeology? Keynote 3 by Jürgen Landauer Let the landscape talk about its archaeology: from AI based site detection to Vision Language Models
16:00-17:00	coffee break and poster session
17:00-18:30	Invited lecture 1 by Elise Fovet Data quality (resolution, precision, accuracy) – what you can expect for archaeological exploration Invited lecture 2 by Nives Doneus and Michael Doneus What you see is what you get: past land use and lidar methodology in the northern Adriatic coastal region Invited lecture 3 by Łukasz Banaszek The cream of the crop! ALS data for cropmark archaeology
19:00-20:00	dinner

Wednesday, 29 October 2025

7:00-8:15	breakfast
8:45-10:00	Introductory discussion about workshops
10:00-10:30	coffee break
10:30-12:30	Workshop session 1
12:30-13:30	lunch
13:30-16:30	Field trip afternoon (2 choices) – a walk or the cave
17:00-19:00	Workshop session 2
19:30-21:00	International dinner

Thursday, 30 October 2025

7:00-8:15	breakfast
8:45-10:45	Workshop session 3
10:45-11:15	coffee break
11:15-12:30	Round table discussion and closing session
12:30-13:30	lunch

Workshops

Gentle introduction in the archaeological interpretation of airborne lidar derived data

Rebecca Bennett, Luka Škerjanec
In this workshop we take a look at the process of identifying and interpreting potential archaeological features from lidar derived data. Using the EAC good practice guidelines as a guide, we will explore interpretation as an iterative process, thinking about the challenges of what to record and how. Participants are encouraged to bring their own examples for discussion.

Archaeological ground point filtering of lidar point clouds

Michael Doneus
An important task in the processing of airborne laser scanning data is the derivation of appropriate models (terrain model, surface model, DFM). In this context, classification and reduction of the point cloud to the required classes (filtering) play an important role and are crucial for usability. Setting up such classification and filtering workflows can be time-consuming and prone to information loss, especially in geographically heterogeneous landscapes.
AFwizard is an open source Python package (https://afwizard.readthedocs.io/en/latest/) designed to improve the productivity of ground point filtering workflows in archaeology and beyond. It provides a Jupyter-based environment for human-in-the-loop, spatially heterogeneous ground point filtering. It will be presented at the workshop.
Participants will receive basic information on the process of archaeological filtering and its pitfalls. Through hands-on training, they will learn how to study the effects of different algorithm and parameter combinations on digital terrain modelling in a practical and time-saving way.

Automatic Detection of Archaeological Features (ADAF)

Žiga Kokalj, Nejc Čož, Jürgen Landauer
The rapid development of image analysis techniques and the increasing availability of high-quality airborne laser scanning data (ALS, lidar) are encouraging the use of machine learning in archaeology. The ADAF tool consists of two Jupyter notebooks, one for training (creating specific machine learning models) and one for automatic recognition of archaeological features. While the training part is time-consuming and requires suitable hardware, the recognition part can be done on any modern laptop. This workshop aims to provide participants with hands-on experience with the software and enable them to use it independently for their own projects. We will cover the basics of deep learning in archaeology, the installation, explain all components of the training part and run the detection process. The currently implemented model is optimised to detect three classes of Irish archaeology (barrows, enclosures, ringforts), but you can also input your own model. The software requires minimal interaction and no prior knowledge of machine learning techniques, which greatly increases its accessibility to the archaeological community. The software is tested to run on Windows.

Processing UAS lidar data

Jitte Wagen, Luka Škerjanec, Elise Fovet, João Fonte
The advent of UAS-based lidar has brought exciting new opportunities to the field of near-surface remote sensing. Of course, it makes it possible to collect lidar data where none is available, but the flexibility of drone deployment also allows control over data collection, creating new opportunities such as optimising timing, flight patterns and so on and so forth. However, the large amounts of data collected by UAS-based lidar pose challenges for data processing. This workshop will cover strategies for dealing with the classification and filtering of high resolution point clouds using software such as LAStools and CloudCompare.

Presentations

KEYNOTE 1
Attrition, palaeo-landscapes, and the patterns of the past in farmed landscapes

Dave Cowley
The impact of increasingly heavily mechanised cultivation on the character of the historic landscape and the survival of archaeological remains is perhaps self-evident. The distribution of such activity is one of the most important factors that impact the survival and visibility of archaeological remains. But of course, such activity, if not as heavily mechanised, has been taking place for millennia. This paper will take some examples from Scotland to consider the role of ALS data in exploring such landscapes, from thinking about palaeo-landscapes to the enduring impact of farming practice and the difficulties of identifying and understanding the farmed surfaces of the past.

KEYNOTE 2
Do we really need drone-derived lidar data in landscape archaeology?

João Fonte
The advent of drone-based lidar has transformed landscape archaeology, providing unprecedented flexibility and high-resolution data. However, the significant increase in data density raises concerns about the efficiency of data handling and processing, and the practical need for such detail. This presentation will present a comparative analysis of drone-based and airborne lidar data from a number of archaeological sites in the Iberian Peninsula, critically examining the trade-offs between data resolution, processing requirements and archaeological insight.

KEYNOTE 3
Let the landscape talk about its archaeology: From AI based site detection to Vision Language Models

Jürgen Landauer
Today, archaeologists have access to high-quality airborne laser scanning (ALS, lidar) data covering an increasing number of regions around the world. Researchers have made spectacular finds, especially under the canopy. But the abundance of data available, sometimes referred to as the ‘big data problem’, contrasts with the limited number of human experts and calls for the search to be automated. Deep learning, a subfield of artificial intelligence (AI), is currently the most promising method for automating the search for previously unknown features in landscape archaeology.
This talk reports on key findings from numerous AI projects since 2018, searching for features ranging from charcoal kilns, barrows and Celtic grave gardens to large or even amorphous types such as hollow roads and hillforts. We focus on projects of provincial and national scale, and present ideas for solutions to the particular challenges posed by their scale.
We then show how, in the not-too-distant future, researchers will be able to literally ask questions of their field of research using Vision Language Models (VLMs). These siblings of ChatGPT promise much easier access to AI, as they no longer require AI training in advance.

INVITED LECTURE 1
Data quality (resolution, precision, accuracy) – what you can expect for archaeological exploration

Elise Fovet
An increasing amount of airborne laser scanning (ALS) data is now publicly available, offering valuable opportunities for archaeological research. Nationwide ALS surveys and archived datasets continue to grow, yet their quality varies significantly. The quality of the data is impacted by a variety of factors, including differences in flight parameters, sensor technology, scanning seasonality, and advancements in data processing.
ALS surveys are conducted for diverse purposes, but not all are appropriate for cultural heritage studies. In the field of archaeology, lidar applications are highly specialised, prioritising micro-relief detection and landscape anomalies, which demand high-quality data. It is therefore vital to understand the limitations, potential applications, and overall suitability of lidar data originally collected for other purposes.
This lecture introduces key concepts for evaluating lidar datasets and explores the criteria that define high-quality ALS data for archaeological research. What makes a dataset suitable for archaeology? Can less precise data, initially acquired for other applications, still be useful?

INVITED LECTURE 2
What you see is what you get: past land use and lidar methodology in the northern Adriatic coastal region

Nives Doneus and Michael Doneus
The widespread use of large-scale archaeological prospection techniques over the last two decades has shifted the archaeological focus from an interest in individual sites to a holistic view of cultural landscapes. In this context, and particularly in the study of Mediterranean landscapes, airborne laser scanning (ALS, lidar) serves as a robust tool that can yield particularly promising results. ALS data offer insights into the relationship between human settlement activities and the natural environment and help to understand how the land has been used to meet human needs. The development of relative stratigraphic information from lidar data is an essential aspect, as it can provide an insight into the temporal development and change of regional land use patterns.
The results of several Croatian case studies, mainly from the Roman period, show that land use remains, such as relics of agro-pastoralism and land division, can be considered as part of the material culture and an expression of culture and society. Decoding past (agricultural) landscapes from lidar data is therefore a viable approach to better understanding of historical processes.

INVITED LECTURE 3
The cream of the crop! ALS data for cropmark archaeology

Łukasz Banaszek
Applications of airborne laser scanning in archaeology tend to focus on examining surface topography. Less well explored is the potential for such data to investigate levelled archaeological features that are expressed through differential arable crop growth (cropmarks). This is partly because topographic and archaeological airborne lidar surveys are often undertaken outside the crop growing season, limiting the potential to explore this aspect of lidar data.
However, ALS sorties can take place at the peak period of cropmark formation if the recording of surface objects, e.g. buildings, trees, and power lines, takes priority over detailed representation of the terrain. In addition, a wet and windy cold season in countries such as Scotland, i.e. November to March, complicates data acquisition, thus limiting the leaf-off data acquisition window. As a result, data collected over spring and summer is available for the analysis of cropmarks.
The lecture evaluates ALS-derived height and radiometric data against traditional aerial reconnaissance outcomes across the Scottish Lowlands. Also drawing on Sentinel-2 satellite imagery and targeted field visits, I discuss the utility of large-area leaf-on airborne laser scanning data in recording cropmark archaeology.