Assessing the Risks of Enhancing the Current Europe’s ADA Web Map

with Ground Movement Classiﬁcation Data

Jos

e A. Navarro

, Anna Barra

and Mar

ıa Cuevas-Gonz

alez

Centre Tecnol

ogic de Telecomunicacions de Catalunya-CERCA, Parc Mediterrani de la Tecnologia – Building B4,

Av. Carl Friedrich Gauss 7, 08860 Castelldefels, Spain

Keywords:

Active Deformation Areas, EGMS, Geoprocessing, Ground Movement Classiﬁcation, Open-Source, Web

Map.

Abstract:

The European Ground Motion Service is offering data on ground movement across Europe with millimetre

precision. With the intention of helping in the interpretation of such a large volume of data, the CTTC has al-

ready created an online Active Deformation Areas (ADA) web map, which can be consulted freely. The CTTC

is considering the possibility of enhancing the said web map by including the causes explaining why ADAs

occur. This article presents the changes in the self-developed ADAtools to make possible such enhancement

and analysis of the impacts on the current implementation of the web map as well as an early assessment of

the risks that such changes would imply.

1 INTRODUCTION

The European Ground Motion Service (EGMS) (Eu-

ropean Environment Agency, 2021b; Crosetto et al.,

2020) has made available to the public the deforma-

tion measurements of practically all of Europe. Three

levels of products are distributed, namely: (a) the ba-

sic one, which consists of line-of-sight (LOS) velocity

maps referred to a local reference point for both as-

cending and descending orbits, (b) the calibrated one,

which is obtained by correcting the data of the ba-

sic product using a model derived from data from the

Global Navigation Satellite Service (GNSS) as a ref-

erence, and ﬁnally, (3) Ortho, which consists of the

horizontal and vertical displacements calculated from

the reference data.

This constitutes a real plethora of information that

allows for continental-scale projects to be undertaken.

However, the interpretation of these data as they are

available is difﬁcult and complex, and it is conve-

nient to use tools that offer a higher level of abstrac-

tion and consequently facilitate their understanding.

The ADAtools have been carrying out this type of

task since 2018; the most notable tasks they are capa-

ble of performing are (a) identifying Active Deforma-

https://orcid.org/0000-0001-7877-1516

https://orcid.org/0000-0001-6254-7931

https://orcid.org/0000-0002-4988-5669

tion Areas (ADAs) (Barra et al., 2017; Navarro et al.,

2020) and (b) classifying said areas according to up

to four different processes (subsidence, constructive

settlement, sinkhole, and landslide).

Initially, the ADAtools were used to process ar-

eas of limited extension—that is, what could be clas-

siﬁed as local or regional projects. Since the ap-

pearance of the EGMS data, it has been possible

to address continental-level targets. One example is

the calculation of ADAs for all of Europe with the

data available between 2015 and 2020, the ﬁrst de-

livery of the named EGMS baseline: the European

ADA web map (Navarro et al., 2022; Navarro et al.,

2023) (see Figure 1); another project, currently on-

going, is aimed at the generation of wide-area differ-

ential deformation maps (Shahbazi et al., 2023) indi-

cating the gradient of the deformation ﬁeld, also from

EGMS basic products and again covering mostly all

Europe—although, at the moment of writing this pa-

per, only a small part of Spain has been published.

These two projects have been materialized as web

maps and WMS / WMF layers. Both web maps may

be accessed at https://groundmotionadas.com/.

The EGMS is committed to update the informa-

tion it delivers on a regular basis; this makes possi-

ble (and advisable) to update, as well, the European

ADA web map, incorporating the new information

available. Since this means that the ADAs for all Eu-

rope should have to be recomputed, this situation may

Navarro, J., Barra, A. and Cuevas-González, M.

Assessing the Risks of Enhancing the Current Europe’s ADA Web Map with Ground Movement Classiﬁcation Data.

DOI: 10.5220/0012720900003696

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 10th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2024), pages 195-202

ISBN: 978-989-758-694-1; ISSN: 2184-500X

195

Figure 1: The web map for the ADAs of all Europe

(https://groundmotionadas.com/).

be seen as an opportunity to reconsider what kind of

ADA data the new version of the web map could in-

clude. The authors believe that the most relevant data

to consider would be the reason why ADAs occur—

that is, what kind of ground movement process is pro-

ducing these.

The next sections will describe what challenges

(in terms of risk) that such changes would imply, con-

sidering the tools to use or develop, the impact on

the current data production workﬂow and hardware

resources.

2 THE TOOLS

The workhorse to produce the ADAs and afterwards

their classiﬁcation is the ADAtools toolset (Barra

et al., 2017; Navarro et al., 2020). The main tools

involved in the process are ADAﬁnder and ADAclas-

siﬁer.

ADAﬁnder takes care of identifying the ADAs. It

takes the Measurement Points (MP) distributed by the

EGMS as input, producing the polygons deﬁning the

said ADAs as well as some data characterizing them,

such as their mean velocity, area, or an assessment

of the quality of the ﬁnding—that is, how probable

is that the area identiﬁed as active be active indeed.

ADAﬁnder is a seasoned, reliable tool, used in many

projects since 2018. Therefore, the authors consider

that it may be used for this new project in its current

state.

ADAclassiﬁer starts where ADAﬁnder ends. It

takes the ADAs produced by ADAﬁnder and runs a

series of tests to determine, for each of these ADAs,

what is the kind of ground movement process that

most probably caused it. Up to four different pro-

cesses were checked by the original version of this

tool: subsidence, constructive settlement, sinkhole,

and landslide. To do it, it implemented a single de-

cision subtree for each of these processes. Figure 2

depicts the one used by the landslide classiﬁcation

Figure 2: Single decision tree for landslides in the old ver-

sion of ADAclassiﬁer.

process. These trees relied on several datasets and

thresholds (denoted as “Th01”, “Th02” and “Th03”

in the said ﬁgure) provided as inputs by the user. The

datasets include a digital terrain model, several kinds

of inventories (for subsidences, constructive settle-

ments, sinkholes, and landslides as well as a geologic

map) and horizontal displacement data. Most of these

inputs are optional, but their absence will make some

of the checks unavailable.

Note that, unlike ADAﬁnder, the original version

of ADAclassiﬁer was not a seasoned tool. In fact, it

had some limitations and problems that prevented the

authors to use it “as is” in this project. The next sec-

tion explains why.

3 IMPROVING ADAclassiﬁer

ADAclassiﬁer, the tool of the ADAtools in charge of

classifying the different phenomena of ground move-

ments, suffered from a series of defects that advised to

improve or replace it before undertaking any massive

data processing campaign if the risk of generating in-

correct data was to be avoided. The most important

defects were the following:

• All tests relied on the consultation of inventories

(such these for landslides or sinkholes). The per-

formance observed when querying the inventories

was far from satisfactory. The reason, a non-

optimal implementation of the polygon intersec-

tion search algorithms. In fact, said performance

was of order n × m, where n and m represent, re-

spectively, the number of polygons to be queried

(ADAs) and the polygons available in the inven-

tory. Since the implementation of the inventory

GISTAM 2024 - 10th International Conference on Geographical Information Systems Theory, Applications and Management

196

software was the same for all inventories the per-

formance problem affected all classiﬁcation pro-

cesses. This meant that processing a single EGMS

burst could take about 4 hours; this might be per-

fectly valid to process a small area, but consider-

ing that there are about 15,000 EGMS bursts, the

time required to ﬁnish the processing of all data

would take about 2,500 days working non-stop.

This problem alone would make the project un-

feasible.

• Decision trees were too simple, due to the early

stage of development of the concept when ADA-

classiﬁer was ﬁrst implemented. The concurrence

of different factors that could inﬂuence the classi-

ﬁcation process was not considered. On the con-

trary and based on a relatively small number of

data and thresholds, an attempt was always made

to reach a conclusive decision regarding the type

of ground movement process that was being ver-

iﬁed. This can be seen in Figure 2, which shows

the unique decision tree for the landslide detection

process in the old version of ADAclassiﬁer. Note

that the trees for the rest of the processes were of

similar simplicity.

• The data formats accepted for certain types of in-

formation were limited or not very widespread.

For example, horizontal displacement data was

accepted in the format generated by los2hv only

(los2hv is another tool included in the ADAtools

that computes the separate horizontal and ver-

tical components plus the total displacement of

the ground displacement measured with Persis-

tent Scatterer Interferometry (PSI) technologies

along the satellite’s line of sight); the digital ter-

rain models could only be used in the ENVI native

format.

• The uplift detection process was not implemented.

The defects mentioned have a clear impact on three

important cornerstones: performance, reliability, and

ﬂexibility of the tool. To address these shortcom-

ings, ADAclassiﬁer has undergone a deep revision

and improvement process. Obviously, the improve-

ments were aimed at eliminating or mitigating the

mentioned defects: the performance of searches in

the inventories now work in logarithmic instead of

quadratic times, so processing a burst takes, usually,

just a few minutes; new formats have been added,

such as the popular GeoTIFF for storing horizontal

and/or vertical displacement data as well as for dig-

ital terrain models. Additionally, the uplift detection

process has been incorporated.

The improvements in the performance of this tool

seems to make possible the processing of all Europe.

The tests carried out seem to point in this direction;

however, the authors are not sure that this perfor-

mance will be maintained uniformly for all available

bursts from the EGMS, since such performance de-

pends on several factors; one of them, for instance, is

the complexity of inventories—tests have shown that

when these include polygons with a very high number

of vertices (apparently, more than 10,000) the perfor-

mance is degraded, due to the increased complexity

of the intersection operations.

Besides performance, the most important change

is the one related to the reliability of the results. The

new version of ADAclassiﬁer now incorporates a set

of decision subtrees, and not just one tree, for each

of the ground movement processes it checks; for in-

stance, the test to decide whether an ADA is a land-

slide consists of four decision sub-trees instead of just

one. In this way, a much larger set of factors that may

have some impact when deciding whether an ADA

corresponds to a certain process can be checked in

a non-exclusive way. Each of these subtrees gener-

ates a score; the scores of each subtree are added,

thus reaching a ﬁnal, total score, that collects much

more information than that provided by the old, orig-

inal ADAclassiﬁer trees. Based on the total score, a

class is assigned to the ADA. The classes are “It is

not X”, “It may be X”, “It should be X” and “It is X”,

where “X” stands for the different kinds of ground

movement processes.

Figure 3 depicts just one of the four decision

subtrees used in the landslide classiﬁcation process.

There, “ThLa04” stands for some threshold input by

the user and the numbers on the leaves stand for the

points scored by each one. Conditions such as “As-

pect in A” or “VLOS is consistent with aspect” sum-

marize some checks that are explained in the applica-

tion’s user guide.

With these changes, the authors believe that the

new ADAclassiﬁer could be a suitable tool for mass

data production. This, however, must be conﬁrmed

in the near future, processing more datasets to gather

more performance data for this tool. Therefore, the

authors perceive a risk here that should be addressed

before taking any steps towards the implementation

of this enhancement.

4 AUTOMATING THE MASS

PRODUCTION PROCESS

The transition from processing a small number of

datasets to handling information on a continental

scale necessitates the identiﬁcation and resolution of a

novel set of challenges before embarking on produc-

Assessing the Risks of Enhancing the Current Europe’s ADA Web Map with Ground Movement Classiﬁcation Data

197

Figure 3: One of the four decision subtrees for landslides in

the new ADAclassiﬁer.

tion activities.

The ﬁrst challenge is automation. It does not make

sense to dedicate human effort to executing the dif-

ferent applications that will integrate the production

workﬂow, dataset by dataset. Therefore, it is conve-

nient to have meta-tools that allow the required tools

to be executed as many times as necessary, reducing

human intervention to the minimum indispensable.

There are already meta-tools to automate the execu-

tion of both ADAﬁnder and ADAclassiﬁer. Conse-

quently, there are no risks related to the automation of

these two tools.

The second challenge is related to the organiza-

tion of information. This includes an important fac-

tor such as the systematization of the ﬁle nomen-

clature; note, however, that this issue is present in

all mass production systems. Without a normal-

ized nomenclature, it is impossible to automate pro-

duction. Fortunately, this problem is not such in

this case, since the original EGMS data source it-

self has systematized the naming of data; if this

nomenclature is adapted with standardized variants

to denote the different types of by-products gener-

ated by the system, the problem is solved. For

example, given some downloaded ﬁle whose name

is “some ﬁle”, the variants “some ﬁle ADA” or

“some ﬁle ADA CLASSIFIED” may be selected to

represent such by-products. This convention was al-

ready adopted for the ﬁrst version of the European

Figure 4: Spatial overlap of EGMS data. Source: (Navarro

et al., 2023).

ADA map; now, it should be extended to consider the

by-products generated by the new ADAclassiﬁer.

More serious is the problem of how input data has

been organized: EGMS has processed the Sentinel-

1 bursts that overlap to guarantee full coverage. As

can be seen in Figure 4, there is an overlap between

contiguous bursts. This causes the ADAﬁnder appli-

cation to generate repeated (although not identical)

ADAs covering the same area in those zones where

such bursts overlap (Figure 5), which, in the speciﬁc

case of this web map, is unacceptable. Therefore, it is

necessary to carry out additional data cleaning tasks

due to how the information has been organized at the

source.

This problem was already detected when imple-

menting the European ADA web map and had to be

solved by implementing the purge overlaps tool. The

biggest problem with this tool is the need to load all

the ADAs in Europe simultaneously, to eliminate the

overlaps at once (Navarro et al., 2023). However, the

servers where this process is carried out are, fortu-

nately, capable of handling the load that this entails;

nonetheless, it is worth to remark the problem here,

since it must be highlighted how the organization of

data may impact in how data must be processed or,

worse, whether it is possible to process it due to ca-

pacity problems.

Considering these two issues in advance (ﬁle nam-

ing and ADA overlapping) and taking the appropri-

ate measures the risk of a project failure due to data

organization problems is, according to the author’s

standpoint, eliminated. Not considering these factors

could, conversely, lead to a high-risk project.

No other risks related to data organization aspects

are foreseen at the moment of writing this paper.

GISTAM 2024 - 10th International Conference on Geographical Information Systems Theory, Applications and Management

198

Figure 5: The need to remove overlaps. Source: (Navarro

et al., 2023).

5 FROM DATA TO THE WEB

Assuming that it is possible to generate the classiﬁca-

tion of all ADAs in Europe using the tools (and meta-

tools for automation) mentioned in the previous sec-

tions, there are still problems to be solved. This sec-

tion explains how data formats are one of these prob-

lems.

The new version of the web map should display

classiﬁed ADA data, which ought to be stored some-

where. The most common way to store data to be used

by web-based applications is a relational database, as

for example PostgreSQL. In fact, this is the solution

adopted by the current web map.

However, ADAclassiﬁer generates ESRI shape-

ﬁles, which cannot be inserted directly into a

database. Therefore, it will be necessary to create

a tool to convert these shapeﬁles into ﬁles contain-

ing SQL (Standard Query Language) statements. By

mean of these ﬁles it will ﬁnally be possible to trans-

fer the information to the database.

A tool like this already exists, ADA2PGIS, but

it should be modiﬁed to handle the new data ﬁelds

to consider—at least, the code representing the most

probable ground movement process causing the ADA.

This would not be so great a change, but it should not

be forgotten. Furthermore, the tool should be able to

automate the conversion process as much as possible,

considering again the issues (time, human resources)

related to the processing of 15,000 datasets. That was

already solved for the current version of the Euro-

pean web map, but the changes in the data production

workﬂow and therefore in the set of by-products to

deal with would imply that the tool should have to be

reimplemented to handle such changes. The authors,

however, believe that the development of the new ver-

sion of ADA2PGIS would not be a challenge, in terms

of risk.

It is not just the conversion tool, however, the only

element that would suffer a change due to the new in-

formation considered. The structure of the database

itself should be changed, in fact, by two reasons: the

ﬁrst one, the number of elements in the ADA’s time

series (new epochs, and thus, new MPs are included

with each EGMS update), the second, the code stat-

ing the reason of the ground movement. The authors

decided that, although mandatory, this change should

not be considered as a risk, since it is present event in

the case the ground movement code is not considered

in future updates, for the database must be modiﬁed

due to the increase in size of the time series when new

updates from the EGMS are implemented in the web

map. In short, the risk of modifying the structure of

the database would be the same no matter the number

of new ﬁelds to include.

6 THE WORKFLOW

Sections 2 to 5 delve into the project’s ADAtools

toolset, exploring both existing tools and those need-

ing development. Additionally, these sections tackle

the most important challenges that must be addressed

and overcome. Here, a global picture is presented ex-

plaining how these tools should work together.

Figure 6 depicts how, starting with the deforma-

tion maps downloaded from EGMS, the information

about ADAs is stored in the database.

In the ﬁgure, green blocks stand for automated

processes, that is those able to process more than one

burst at a time—ideally, all the bursts at once. Note

that the process to purge the repeated ADAs lying

in overlapped areas is run just once due to the con-

straints of the problem. On the other side, download-

ing the deformation maps from the EGMS is a manual

process that must be repeated until all the necessary

information has been retrieved. It is expected, how-

ever, that improvements on the EGMS Explorer (the

interface to download EGMS data, (European Envi-

Assessing the Risks of Enhancing the Current Europe’s ADA Web Map with Ground Movement Classiﬁcation Data

199

Figure 6: The candidate workﬂow: from EGMS MPs to

classiﬁed ADAs in a database.

ronment Agency, 2021a)) will make possible the au-

tomation of these downloads.

The ﬁrst three steps in Figure 6 would be manda-

tory even if no enhancements to the web were made,

since incorporating the latest data updates to it can-

not be done without them. The greatest risk related

to the changes in the workﬂow is the potential, still

unknown increase of processing time due to the clas-

siﬁcation of the ADAs using ADAclassiﬁer (see sec-

tion 3); at the moment of writing this paper, not

enough tests had been performed to assess it with a

comfortable level of conﬁdence.

7 THE WEB MAP

Drawing on the past experience from the implemen-

tation of the currently published version of the web

map, the authors foresee no signiﬁcant challenges to

enhance it to show the new data about ADAs. Basi-

cally, changes are related to the way relevant informa-

tion would be shown.

The goal would be visually displaying as much

key data as possible without overwhelming users.

While ADA’s velocity was the primary focus in

the initial map (represented by colour-coded painted

ADAs), now there would be two key elements: veloc-

ity (as in the current version) and ground movement

processes (the new data). Ideally, both should be visu-

ally integrated to avoid confusion and unlock deeper

insights.

The authors envision a solution using distinct

colours for the perimeter and inner area of each ADA,

as illustrated in Figure 7. This approach would lever-

age colour coding effectively: the perimeter colour

would represent velocity, while the inner area would

Figure 7: ADA colour codes for velocity (perimeter) and

ground movement process (interior).

depict the ground movement process. This conven-

tion would serve to enhance clarity and avoiding in-

formation overload, potentially improving user com-

prehension; this convention could be reversed (ex-

changing the roles or perimeters and inner areas).

This double colour scheme could be easily imple-

mented using the styles included in GeoServer (see

section 8); therefore, no problems are foreseen in this

regard. The authors are conﬁdent in the technical fea-

sibility of this approach.

8 THE SERVER

To make the web map available to the public, one or

more servers are needed, which should host the appli-

cations themselves (the web map) as well as the soft-

ware stack necessary to implement the service. Like-

wise, the data must also reside on one of these servers.

The server used with the current version of

the web map is not affected by performance is-

sues (Navarro et al., 2022). Consequently, the authors

are truly conﬁdent that it would be perfectly valid to

host the new version. Nowadays, information about

the amount of data to store in the database as well as

the workload that the server must support is available.

The inclusion of the new update from the EGMS and,

eventually, that of the ground movement information,

would imply a rather negligible increase of both space

or performance requirements. Considering that server

is working comfortably with the current workload,

there is a wide margin of manoeuvrer to increase the

volume of data or performance requirements, so no

problems nor risks are foreseen regarding this issue.

The software stack and operating system of the

server would remain unchanged:

GISTAM 2024 - 10th International Conference on Geographical Information Systems Theory, Applications and Management

200

• PostgreSQL (The PostgreSQL Global Develop-

ment Group, 2024) as the database manager, in-

cluding PostGIS (PostGIS Project Steering Com-

mittee, 2024) to store, index, and query geospatial

data. These will keep all the information relative

to ADAs.

• GeoServer (OSGEO, 2024), for sharing spatial

data. Used as a bridge between the database and

the web map. Transforms raw database data into

WMS / WMF layers.

• Apache HTTPD (The Apache Software Founda-

tion, 2024b) and Tomcat (The Apache Software

Foundation, 2024a) servers. These are technol-

ogy enablers. They make possible (1) to run

GeoServer and (2) to have a web server to host

the web map itself.

• The web map application. Relies on GeoServer

to retrieve ADA data and on public base maps

such as OpenStreetMap (OpenStreetMap Founda-

tion, 2024) to add a background cartography layer.

This is the human interface.

• The operating system of the physical server host-

ing the previous software components is Ubuntu

Linux server edition (Canonical Ltd., 2024).

This selection of technologies incorporates only free

and open-source software components. This has an

important economic impact on the implementation of

the project, as there is no cost to use them. Conse-

quently, the adoption of this software stack reduces

the risk of the project. Figure 8 shows the relation-

ships between all the components making the system.

9 CONCLUSIONS

With this paper the authors are trying to assess the

risks related to the addition of a new feature in an al-

ready working web map. This enhancement has been

discussed in the context of a necessary system update

due to the emergence of new EGMS data.

Several aspects have been discussed, such as the

required reliability and performance of the tools in-

tervening in a mass production process, the need for

automation, the unexpected problems that data orga-

nization may produce, the change in data formats or

database structure or the need of a server to imple-

ment the said system. These subjects have been ex-

plored under the light of risk assessment, trying to

identify the weakest link or links in the chain, so a de-

cision might be taken based on solid (or more solid)

evidence.

The experience developing and implementing the

original European ADA web map has proven es-

Figure 8: The architecture of the server.

sential to assess the risks of almost all the points

discussed—as, for instance, how difﬁcult would be to

adapt ADA2PGIS to convert from ESRI shapeﬁles to

SQL or whether the selected server and software stack

would be enough to implement the whole system.

The most difﬁcult element to assess is the new ver-

sion of the ADAclassiﬁer tool. Its reliability has been

thoroughly tested, so the problem is related only to

performance. Although it is much faster than the orig-

inal version, not enough tests have been yet carried

out to be reasonably sure that a mass production cam-

paign may be started. The authors have noticed that

such performance might be inﬂuenced by the com-

plexity of the polygons in the inventories, so more

tests are mandatory before any further steps are taken.

On the other side, the intersection of ADAs and

inventories are just a subset of the tests (subtrees, see

section 3) that ADAclassiﬁer carries out. In fact, in-

ventories are optional inputs to ADAclassiﬁer; when

not present, the checks (subtrees) related to inven-

tories are simply ignored. This, combined with the

fact that, usually, inventories are not available, might

make the ADAclassiﬁer performance issue less im-

portant. It is still unknown which inventories (for

each kind of ground movement process) will be avail-

able if it is decided to process all of Europe. This, to-

gether with the performance of ADAclassiﬁer, is the

second unknown to be solved.

In conclusion, the analysis presented herein al-

lows the authors to afﬁrm that the enhancement of the

web map is possible from the technical standpoint;

no risk is perceived on this side. On the contrary,

Assessing the Risks of Enhancing the Current Europe’s ADA Web Map with Ground Movement Classiﬁcation Data

201

there are serious doubts from the temporal standpoint:

the lack of solid knowledge about the performance of

ADAclassiﬁer as well and about the availability of the

inventories, makes very difﬁcult to assess how much

time would be needed to complete the whole project,

and whether such amount of time is affordable. Fur-

ther work on this direction is therefore needed.

Note, however, that the current version of the

ADAtools is reliable and performant enough as to be

used in local / regional projects. It is available for free

and may be obtained contacting any or the authors.

ACKNOWLEDGEMENTS

This work has been developed in the framework of

project SARAI (PID2020-116540RBC22), funded by

(MCIN/ AEI /10.13039/501100011033).

REFERENCES

Barra, A., Solari, L., B

ejar-Pizarro, M., Monserrat, O.,

Bianchini, S., Herrera, G., Crosetto, M., Sarro, R.,

Gonz

alez-Alonso, E., Mateos, R. M., Lig

uerzana, S.,

opez, C., and Moretti, S. (2017). A Methodology to

Detect and Update Active Deformation Areas Based

on Sentinel-1 SAR Images. Remote Sensing,, 9(10).

Canonical Ltd. (2024). Ubuntu server - for scale out work-

loads | ubuntu. https://ubuntu.com/server. Accessed:

2024-03-14.

Crosetto, M., Solari, L., Mr

oz, M., Balasis-Levinsen, J.,

Casagli, N., Frei, M., Oyen, A., Moldestadk, D. A.,

Bateson, L., Guerrieri, L., Comerci, V., and Andersen,

H. S. (2020). The Evolution of Wide-Area DInSAR:

From Regional and National Services to the European

Ground Motion Service. Remote Sensing, 12:2043.

European Environment Agency (2021a). EGMS. https://

egms.land.copernicus.eu/. Accessed: 2024-03-14.

European Environment Agency (2021b). European

Ground Motion Service – Copernicus Land Monitor-

ing Service. https://land.copernicus.eu/pan-european/

european-ground-motion-service. Accessed: 2024-

03-14.

OSGEO (2024). GeoServer. https://geoserver.org/. Ac-

cessed: 2024-03-14.

Navarro, J. A., Garc

ıa, D., and Crosetto, M. (2022). An

in-house, self-developed low-cost WebGIS relying on

open-source and self-made tools to show the active

deformation areas across Europe. In Proceedings of

the 8th International Conference on Geographical In-

formation Systems Theory, Applications and Manage-

ment (GISTAM) - Volume 1, pages 99–106. INSTICC,

SciTePress.

Navarro, J. A., Garc

ıa, D., and Crosetto, M. (2023).

New challenges in the implementation and exploita-

tion of a low-cost web map of the active deforma-

tion areas across Europe. In Proceedings of the 9th

International Conference on Geographical Informa-

tion Systems Theory, Applications and Management

(GISTAM) - Volume 1, pages 192–199. INSTICC,

SciTePress.

Navarro, J. A., Tom

as, R., Barra, A., Pag

an, J. I., Reyes-

Carmona, C., Solari, L., Vinielles, J. L., Falco, S., and

Crosetto, M. (2020). ADAtools: Automatic Detec-

tion and Classiﬁcation of Active Deformation Areas

from PSI Displacement Maps. ISPRS Int. J. Geo-Inf.,

9:584.

OpenStreetMap Foundation (2024). OpenStreetMap. https:

//www.openstreetmap.org/about. Accessed: 2024-03-

14.

PostGIS Project Steering Committee (2024). PostGIS

— Spatial and Geographic Objects for postgreSQL.

https://postgis.net/. Accessed: 2024-03-14.

Shahbazi, S., Navarro, J. A., and Barra, A. (2023). Com-

puting and sharing the differential deformation of the

ground at a continental level using public EGMS data.

In Proceedings of IV Conference on Geomatics Engi-

neering. MDPI.

The Apache Software Foundation (2024a). Apache

Tomcat

– Welcome! http://tomcat.apache.org/. Ac-

cessed: 2024-03-14.

The Apache Software Foundation (2024b). Welcome!

– The Apache HTTP Server Project. https://httpd.

apache.org/. Accessed: 2024-03-14.

The PostgreSQL Global Development Group (2024). Post-

greSQL: The world’s most advanced open source

database. https://www.postgresql.org/. Accessed:

2024-03-14.

GISTAM 2024 - 10th International Conference on Geographical Information Systems Theory, Applications and Management

202