MAPPING LAVA

FLOW AND

ASSESSING DAMAGE:

VOLCÁN CUMBRE VIEJA, LA PALMA

Robert Kozub

GISC 6325

Fall 2021

image: Michael Risch

Overview

La Palma is the north-west island of the

Canary Islands, Spain. The total

population as of 2021 was 86,267

(source: statista).

On September 19th, Volcán Cumbre Vieja

began to erupt, resulting in over 1000

buildings and roads destroyed and 1

known death.

In addition to the destruction caused by

the eruption, as many as 6,400 people

have been forced to evacuate the area

(per science.org).

La Palma

North Africa

La Palma, a part of the Canary Island, is a

Spanish Territory located off the coast of

the Western Sahara in North Africa.

Methods

Landsat 8 data for September 26 th & October

28th , while Sentinel 2 was used for October

15th due to cloud coverage visibility.

Using ERDAS Image, the layers were stacked for

bands 1-7. The images were then evaluated using

various band combinations to best see where

lava was currently flowing and had already been.

Bands 7,6,5 were used for Landsat 8 images.

Images were then evaluated in ArcGIS Pro and

the estimated flow was digitized. Building and

transportation data were found on Coperncus

(European Emergency Management Service).

Land Surface Brightness Temperature was

calculated using band 10 and bands 4&5 (for

NDVI). The purpose of calculating this is to see

what the temperature of the active lava flow vs

resided.

Purpose

Count buildings and roadway segments

were within the lava flow polygons that

were digitized analyzing remotely sensed

imagery.

Lava Flow:

9/26/2021

This image represents the

approximate lava flow area

on 9/26/2021.

The imagery used was

Landsat 8.

As you can see, there is

smoke stemming from the

eruption point (red dot).

This is nearly a week after

the volcano erupted and

has amassed an

approximate area of 1.21

sq mi.

Lava Flow:

10/15/2021

This image represents the

approximate lava flow on

10/15/2021, approximately

3 weeks af ter eruption.

Around 3.3 sq mi were

encompassed with lava at

this point in time.

Sentinel 2 imagery was

used for mapping this

polygon as the Landsat

imagery for this data range

contained too much cloud

cover.

Lava Flow:

10/28/2021

This image represents the

approximate lava flow on

10/28/2021, approximately

5 weeks after eruption.

Around 4.13 sq mi were

encompassed with lava at

this point in time.

Landsat 8 imagery was

used here to determine lava

flow.

Land Surface

Brightness

Temperature

Land Surface Brightness Temperature was

calculated using the following method in

QGIS:

1. NDVI Using ERDAS Imagine

(Unsupervised)

2. Calculation of Top Atmospheric Spectral

Radiance

TOA (L) = ML* Qcal + AL

3. Top Atmospheric Spectral Radiance to

Brightness Temperature Conversion

BT = (K2/ (ln (K1/ L) + 1)) − 273.15

Calculation of Top Atmospheric

Spectral Radiance

Top Atmospheric Spectral Radiance to

Brightness Temperature Conversion

The areas in yellow identify

areas where lava flow has

been and blue most likely

shows active lava flows.

Values ranged from 46 to

94 degrees Celsius,

however, active lava has a

much higher temperature.

Data was most likely

skewed due to cloud

coverage and pixel size.

Building

and Road

Damage

Assessment

Building and road data was acquired from

Coperncus, the European Emergency

Management Service. Point layers and line

segments were counted using ArcGIS Pro

within the 10/28/2021 polygon layer.

Of the building points, 1,342 were within

the lava flow area, and can assume they

were either damaged or destroyed.

Of the road line segments, 1,587 were

within the lava flow area, and were either

damaged or destroyed.

Composite Map: Lava Flow

Areas, Damaged/Undamaged

Buildings, & Damaged Roads

649

1160

1342

741

1445

1587

NLumber of Buildings/Roadway Segments

Date

Damaged Buildings and Roads from

9/26/2021 - 10/28/2021

Buildings Roadway Segments

Conclusions

Due to cloud coverage and smoke, the polygon

areas are not as precise as similar maps found

online that most likely used drone imagery,

eliminating most of the previously mentioned

hindrances. The image quality also made if

difficult to determine areas that lava had

previously been opposed to the active flowing

areas, which were more easily visible using

correct band combinations.

Calculating land surface brightness temperature

also proved to be a useful way to identify areas

where lava had been and is actively flowing and

digitization methods could be done using the

temperature output.

In the future, it would be preferable to use drone

imagery as it is able to dodge most cloud

coverage and smoke for more detailed

assessment.

References

•Ea rthEx plo rer. ( 2021). Retrieved 6 Nove mber 20 21, from

h tt ps : // e a rt h ex pl o re r . u s gs . g ov/

•How to calc ula t e Land Surface Temperat ure w it h L andsa t 8

satellite ima ges –G IS Crack. (2 021). Re trieved 6 November

2021, from htt ps:/ / g iscrack.co m/ h ow-to-c alculate - l a nd-

surfa ce-te m pe ra ture - with- l ands at- 8-im a ge s /

•Co perni c u s E merge ncy Manage me nt Serv ic e - Mapping.

(2 021 ). Retriev ed 6 Novembe r 20 21, from

h tt ps : // e me r ge n c y. c o pe r n ic u s .e u / ma p p i ng /

•Scie nce | AAAS. (2021). R etr ieved 6 November 202 1, from

h tt ps : // w w w. s c ie n c e .o rg /

•Statista - The Stat is ti c s Por ta l. (2021). Re trieve d 6

December 20 21, from https :// w ww.s tatis t a.c om/

•., S. ( 201 6). ESTIM AT I ON OF L AND SU RFACE TEM PER ATUR E

OF HIG H RANG E MOUNTA IN LANDSCA P E OF DE VIKULA M

TALUK USING L ANDSAT 8 DATA. Intern atio nal Journal Of

Rese arch In Engineering A nd Te chnology,05(01), 9 2 -96 . doi:

10.15623/ijret.2016.0501017