Remote Sensing Applications: Society and Environment 32 (2023) 101013

Contents lists available at ScienceDirect

Remote Sensing Applications: Society and

Environment
journal homepage: www.elsevier.com/locate/rsase

Using PlanetScope images to investigate the evolution of small

glaciers in the Alps
Giulia Tarca a, Martin Hoelzle b, Mauro Guglielmin a, *
a Department of Theoretical and Applied Sciences, Insubria University, Via Dunant, 3, 21100, Varese, Italy
b Department of Geosciences, University of Fribourg, Chem. du Musée 4, 1700, Fribourg, Switzerland

ARTICLE INFO ABSTRACT

Keywords: Small glaciers (having an area smaller than 1 km2) dominate in terms of total number in the Alps
Small glaciers and it is important to study their evolution. In this study we assessed the potential of daily, high
Remote sensing resolution (3 m pixel size) PlanetScope images for investigating small Alpine glaciers. We com-
PlanetScope pared the evolution of five small glaciers, covering an area lower or slightly higher than 1 km2,
Alps
located in different areas of the Alps and having different characteristics and meteorological con-
Climate
ditions. The investigated glaciers are: Bors (0.645 km2, Piedmont, Italy), Seewjinen (1.339 km2,
Valais, Switzerland), Pizzo Ferrè (0.442 km2, Lombardy, Italy), Gran Zebrù (0.579 km2, Lom-
bardy, Italy) and Solda Orientale (0.829 km2, South Tirol, Italy). We manually mapped the glac-
ier outline and the highest position of the transient snowline (TSL) through PlanetScope images
from the end of the ablation season for each year from 2017 to 2021 and each glacier. The snow
cover area ratio (SCAR) was retrieved using the TSLs.
We found that the glaciers lost between 4 and 14% of their area between 2017 of 2021, with a
mean area loss of −1.5% per year. The strongest shrinkage (14%) is recorded at the smallest glac-
ier (Pizzo Ferrè). The highest TSL showed a variability from year to year. For two glaciers, a su-
pervised classification of the surface facies (snow, ice and supraglacial debris) was performed on
all the available cloud-free images from the 2017–2021 ablation seasons.
PlanetScope images, with their high spatial and temporal resolution proved to be a valuable
tool for studying small glaciers. They permitted to map small glaciers with small uncertainties
(between 2.2 and 5.1% of the glacier area), much lower than with medium resolution satellite
data (e.g. Landsat 15/30 m, Sentinel 10 m) and to increase the number of images available for
analysis. The daily repeat cycle is well suited for increasing the probability to find images from
the end of the ablation season with minimum snow cover, in order to identify the real end-of-
summer snowline.

1. Introduction
In the Alps small glaciers dominate in terms of total number, indeed 92% of the glaciers have an area smaller than 1 km2 (Paul et
al., 2020). Although their total area and volume is small, small glaciers are relevant, since they affect local hydrology, landscape for-
mation and sea-level rise (Huss and Fischer, 2016). However, despite their relevance, Alpine glacier research focuses mainly on
medium and large glaciers (Huss and Fischer, 2016).

* Corresponding author.
E-mail address: mauro.guglielmin@uninsubria.it (M. Guglielmin).

https://doi.org/10.1016/j.rsase.2023.101013
Received 6 February 2022; Received in revised form 30 May 2023; Accepted 14 June 2023
Available online 17 June 2023
2352-9385/© 2023 Elsevier B.V. All rights reserved.
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

The PlanetScope satellite constellation is able to acquire images of the entire land surface of the Earth every day, at a spatial reso-
lution of 3 m (Planet, 2021), offering the possibility to study glaciers at higher spatial and temporal resolution than other commonly
used satellites (e.g. Landsat, Sentinel) and could therefore be suitable for studying small glaciers. PlanetScope images have 4 bands
(Blue, Green, Red, Near Infrared (NIR)) and are orthorectified, geometrically-, radiometrically- and atmospherically-corrected
(Planet, 2021). Their almost-daily availability increases the amount of available data for studying glaciers, especially important in re-
gions that are often cloud-covered. Their high spatial resolution can be useful for a more accurate interpretation of glacier outlines
and of glacier surface facies. Some disadvantages of the PlanetScope images are the absence of the Short Wave Infrared (SWIR) band,
the small region covered by an individual scene (frame size 20–25 km x 12–16 km approximatively (Planet, 2021)) and the limited
freely available quota (5000 km2/month for students and faculty), making them not suitable for large scale studies.
PlanetScope satellites are relatively recent, since the first launch was in 2016, but they have been already used for different appli-
cations in glacier research. PlanetScope images have been used to map glaciers outlines, for example by Pudełko et al. (2018) on glac-
iers having surface areas between 3 and 14 km2 at King George Island (Antarctica), by Chand et al. (2020), for the Parvati glacier
(24 km2) in Himalaya, or by Shaw et al. (2021) for glaciers having surface areas between 3 and 26 km2 in the central Chilean Andes.
PlanetScope images have been also used to map surface velocities of glaciers, for example by Millan et al. (2019) in the southern
Alps of New Zealand, in the Andes and in the European Alps, and by Steiner et al. (2018) at Khurdopin glacier (Pakistan). PlanetScope
images have been used also to generate digital elevation models (DEMs) of glaciers, as Ghuffar (2018) or to detect elevation changes
at successive times at the Khurdopin surging glacier (Pakistan) and to calculate the elevation difference caused by the surge of Shisper
glacier (Pakistan, Aati and Avouac, 2020)). PlanetScope images have also been used to study giant ice avalanches from two glaciers in
Tibet by Kääb et al. (2018) and to map glacier lakes in the Hindu Kush, Karakoram and Himalaya region by Qayyum et al. (2020). Al-
though they have been used for a wide range of applications, at our knowledge PlanetScope images haven't been used to study small
glaciers (<1 km2). Therefore, the aim of this work is to assess the potential of PlanetScope images for mapping small glaciers at an-
nual time scale and investigating their surface evolution during the ablation season. To do so, we selected 5 different alpine small
glaciers, all with a comparable size but different characteristics (aspect, slope, elevation …) and located in different areas of the Alps
with different local meteorological conditions. For each glacier we wanted to map the glacier outline and the highest position of the
transient snowline at the end of the ablation period for each year from 2017 to 2021, and to retrieve the snow cover area ratio, com-
paring the results to the snow covered area obtained from supervised image classification.

2. Study regions
Five glaciers located in different areas of the Alps were chosen, covering an area lower or slightly higher than 1 km2. The glaciers
are, from west to east: Bors (Piedmont, Italy), Seewjinen (Valais, Switzerland), Pizzo Ferrè (Lombardy, Italy), Gran Zebrù (Lombardy,
Italy) and Solda Orientale (South Tirol, Italy) (Fig. 1, Table 1). Bors glacier and Seewjinen glacier belong to the Monte Rosa group,
Pizzo Ferrè glacier to the Tambò-Stella group and Gran Zebrù glacier and Solda Orientale glacier to the Ortles-Cevedale group (Table
1). Bors glacier covers an area of 0.645 km2 (surveyed in 2017), Seewjinen 1.339 km2 (surveyed in 2015), Pizzo Ferrè 0.442 km2 (sur-
veyed in 2016), Gran Zebrù 0.579 km2 (surveyed in 2016) and Solda Orientale 0.829 km2 (surveyed in 2016) (Paul et al., 2019)
(Table 1).
The climate of the Alps is characterized by a high complexity, caused by the interaction between the topography and the general
circulation, and by the coexistence of four different climatological regimes: mediterranean, continental, atlantic and polar (Beniston,
2005). Mean annual precipitation in the Alps show a wide range of values, from about 400 to more than 3000 mm per year (Isotta et
al., 2014). At the scale of the entire mountain range, the distribution of precipitation is characterized by a wet anomaly elongated
along the northern rim of the Alps (about 600 km long), two major wet zones located south of the main crest (one over Lago Maggiore
and one over the Julian Alps) and comparatively dry conditions at several inner-Alpine regions (Frei and Schär, 1998). The Greater
Alpine Region experienced a temperature increase of 1.2 °C in the 20th century, rising about twice as much as the global mean (Auer
et al., 2007). Glaciers in the Alps have been retreating since the end of the Little Ice Age (e.g. Zemp, 2006) and are undergoing a rapid
shrinkage and downwasting, with an area loss of ∼39 ± 9 km2 a−1 (∼1.8% a−1), regionally variable ice thickness changes (−0.5 to
−0.9 m a−1) and a mass loss of 1.3 ± 0.2 Gt a−1 between 2000 and 2014 (Sommer et al., 2020).
Analysing temperature and precipitation data recorded at 7 AWS located nearby the glaciers (Fig. 1), the extrapolated tempera-
tures to the median elevation of the glaciers show mean temperatures between −2.1 and −5.9 °C at the different glaciers in the entire
period (October 2016 to September 2021), while the mean summer temperatures (JJA = June, July, August) range between 2.3 and
6.0 °C (Table 1). Precipitation measured at the closest available stations show that the highest precipitation occurred at Alagna
Miniere (Bors) (average annual precipitation of 1288 mm), followed by Madesimo Spluga (Pizzo Ferrè) (1104 mm), Solda (Solda Ori-
entale) (932 mm), Santa Caterina Valfurva (Gran Zebrù) (875 mm), while the lowest at Passo del Moro (Seewjinen) (760 mm). Con-
sidering the period from October to May (accumulation season), precipitation have the same pattern, with a maximum at Alagna
Miniere (Bors) (average 913 mm) and a minimum at Passo del Moro (397 mm) (Seewjinen) (Table 1).

3. Data and methods

3.1. Data
3.1.1. Satellite images
The PlanetScope constellation is composed by more than 180 satellites, each satellite is a CubeSat 3U form factor (10 cm by 10 cm
by 30 cm). PlanetScope 4-band (Blue: 455–515 nm, Green: 500–590 nm, Red: 590–670 nm, NIR: 780–860 nm) images were used, the

2
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Fig. 1. Geographic location of the five glaciers investigated (white points): 1) Bors glacier; 2) Seewjinen glacier; 3) Pizzo Ferrè glacier; 4) Gran Zebrù glacier; 5) Solda
Orientale glacier. Pink points represent the location of the automatic weather stations (AWS) used in the study: 1) Alagna Miniere and Bocchetta delle Pisse; 2) Passo
del Moro; 3) Madesimo Spluga; 4) Santa Caterina Valfurva; 5) Solda and Solda Madriccio. Black lines represent the 2021 glacier outlines. Red lines represent contour
lines. Yellow lines represent national boundaries. Basemap: Esri World Imagery. (For interpretation of the references to colour in this figure legend, the reader is re-
ferred to the Web version of this article.)

Table 1
Name of the glaciers, Country, Region, Mountain Group, glacier area (km2) (from Paul et al., 2019), year of survey, mean summer temperature (JJA = June, July,
August) (°C) and precipitation during the accumulation season (October–May) (mm), median elevation of the glaciers in 2021 (m a.s.l.).

Glacier Country Region Mountain Glacier Area Survey Mean Summer Precipitation Accumulation Median elevation
Group (km2) year Temperature (JJA) (°C) Season (Oct–May) (mm) 2021 (m)

Bors Italy Piedmont Monte Rosa 0.645 2017 2.8 913 3411
Seewjinen Switzerland Valais Monte Rosa 1.339 2015 2.3 397 3011
Pizzo Ferrè Italy Lombardy Tambò- 0.442 2016 6.0 525 2831
Stella
Gran Zebrù Italy Lombardy Ortles- 0.579 2016 3.6 410 3190
Cevedale
Solda Italy South Ortles- 0.829 2016 5.7 469 2985
Orientale Tirol Cevedale

level 3B PlanetScope Ortho Scene Surface Reflectance data, that are orthorectified, geometrically-, radiometrically- and atmospheri-
cally-corrected (Planet, 2021). The images have a spatial resolution of 3 m and daily temporal resolution. Cloud free images from the
ablation seasons between 2017 and 2021 were downloaded for each glacier, since the first year with available images in these study
sites is 2017.

3.1.2. Digital elevation models

We downloaded the available Digital Terrain Model (DTM) or Digital Surface Model (DSM) of the glaciers and their surroundings.
For Gran Zebrù and Pizzo Ferrè glaciers a regional DTM with 5 m resolution from 2015 is available (vertical accuracy 2 m) (source:
Geoportale della Lombardia https://www.geoportale.regione.lombardia.it/), for Solda Orientale Glacier a DSM (DSM - Ghiacciai
2016/2017) with 0.5 m resolution from 2016 (vertical accuracy 0.1 m) (source: Geoportale Alto Adige http://
geocatalogo.retecivica.bz.it/geokatalog/#!), for Bors glacier a DTM with 5 m resolution (vertical accuracy 0.3–0.6 m) (source: Geo-
portale Piemonte https://www.geoportale.piemonte.it/) and for Seewjinengletscher a DTM (swissALTI3D) with 2 m resolution from
2015 (vertical accuracy 1–3 m) (source: Federal Office of Topography swisstopo https://www.swisstopo.admin.ch/).

3
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

3.2. Methods
3.2.1. Mapping glacier outlines
The glacier outlines of the five glaciers for each year (from 2017 to 2021) were manually delineated on the satellite images. For
each year we chose a cloud-free image from the end of the ablation season, with the minimum seasonal snow (Paul et al., 2013). All
images were acquired between 23/08 and 30/09. To reduce inconsistencies, the outlines were digitized by the same surveyor and on
the same type of data (Racoviteanu et al., 2009). In case of debris-covered ice, we used very high-resolution images (e.g. Google
Earth, ESRI basemaps, aerial orthoimagery) to guide decision on boundaries delineation, as suggested by Paul et al. (2017). More-
over, at the two glaciers where the debris cover is more extended (Gran Zebrù and Solda Orientale glaciers), also field observations
were performed. The eastern tongue of Gran Zebrù is well known by the authors as different works were carried out (Tarca &
Guglielmin, 2022a, 2022b), one including manual measures of debris thickness and GPS data (see Tarca and Guglielmin, 2022a). At
Solda Orientale observations about the presence of ice under debris were carried out.
For each glacier we obtained the glacier area (km2) in the different years. To calculate area uncertainties, we considered a buffer
of ±1 pixel. In the literature different values for area uncertainty are reported, for example: ±1/2 pixel (Ragettli et al., 2016), ±1/2
pixel for clean-ice and ±1 pixel for debris-cover (Mölg et al., 2018), ±1 pixel for clean-ice and ±2 pixel for debris-cover (Paul et al.,
2020). We decided to use an intermediate value of 1 pixel in order to avoid complex calculations.

3.2.2. Transient snowlines

The highest position of the transient snowline at the end of the ablation period was digitized for each year on the PlanetScope im-
ages. The lines were rasterized (5 m cell size) and the average elevation was extracted along each line from the digital elevation mod-
els (e.g. Racoviteanu et al., 2019).
In cases that at the end of the ablation season only small and sparse snow patches were present, a snowline could not be drawn
(e.g. at Gran Zebrù in 2017 and 2018) (Fig. 1 Supplementary).
The TSLs were used to clip the glacier area of the correspondent year in order to obtain the snow cover area ratio (SCAR).
The uncertainty of the elevation of the mapped snowline depends on both the uncertainties of the (1) snowline delineation and (2)
the DEM. The accuracies of the different DEMs are between 0.1 and 3 m (see Section 3.1.2.). Barandun et al. (2015) estimated the un-
certainty of the snowline location as 5 pixels in a horizontal plane on the remote-sensing images, that in our case correspond to 15 m.
It results in vertical errors of 3–4 m at the different glaciers using their mean slopes. Summing it to the uncertainty of the DEM, the re-
sulting total uncertainty of all the snowlines varies between 4 and 6 m for the different glaciers, with a resulting mean uncertainty of
±5 m, that is taken as the uncertainty of the snowlines. The manual delineation of the transient snowlines, based on visual separation
of bare-ice areas from snow-covered sections of the glacier is considered a robust approach, as it integrates the knowledge of the ob-
server on snow-cover depletion patterns (Huss et al., 2013). However, the visual interpretation surely introduces some subjectivity,
that can lead also to errors due to misinterpretation.

3.2.3. Satellite images classification

For the glacier surface facies classification only two of the glaciers were chosen, namely Gran Zebrù glacier and Seewjinen glacier,
the first located in the Central Alps and southern-exposed, the second in the Western Alps and northern-exposed. The classification of
the glaciers surface was performed on all the available cloud-free images from the 2017–2021 ablation seasons using a maximum like-
lihood supervised classification (e.g. Azzoni et al., 2018) of the surface reflectance data with the software ArcMap 10.6. Three classes
were mapped: snow, supraglacial debris and ice. In case of saturation of the visible bands (1, 2 or 3), we performed the classification
on the band 4 if possible, usually not affected by the problem. For Seewjinen glacier, an upper area of the glacier was covered by
shadow in most of the images and therefore was removed to avoid misclassification. In fact, the classifier is not able to discriminate
the different surface types when covered by shadow.

4. Results
4.1. Glacier area
Glacier mapping on the PlanetScope images allowed us to retrieve the surface area of the five glaciers for each year from 2017 to
2021 (Fig. 2, Fig. 3). All the glaciers underwent an area decrease in the considered period, with reductions between 0.026 and
0.068 km2 (Table 2). In terms of percentage, Gran Zebrù glacier showed an area decrease of 4%, two of them, Seewjinen and Solda
Orientale glaciers of 5%, Bors glacier of 8% and Pizzo Ferrè glacier of 14% of the initial area (Fig. 3, Table 2). The glacier that showed
the strongest area reduction is the smallest one (area of 0.445 km2 in 2017).
Considering all the glaciers, the year when the highest glacier area reduction occurred was 2018 (−0.091 km2), while the lowest
occurred in 2020 (−0.046 km2).

4.2. Transient snowlines

The five glaciers showed different highest TSL elevations in the study period (Fig. 4, Table 3). At Gran Zebrù glacier in 2017 and
2018 almost no snow from the current year was present at the end of the ablation season, with only few sparse snow patches. There-
fore, a snowline was not drawn for these two years (Figs. 4d and 1 Supplementary). For each glacier, the snow cover area ratio (SCAR)
was retrieved (Table 3), showing that on average the glaciers with the higher SCAR at the end of the ablation season are Bors, Seewji-
nen and Solda Orientale glaciers (with a mean of 34%, 34% and 35% respectively), while a generally smaller SCAR is found at Pizzo
Ferrè and Gran Zebrù glaciers (8% and 16% respectively).

4
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Fig. 2. Surface area covered by the glaciers from 2017 to 2021: a) Bors glacier; b) Seewjinen glacier; c) Pizzo Ferrè glacier; d) Gran Zebrù glacier; e) Solda Orientale
glacier. Background images are PlanetScope images (acquired 23 to August 30, 2017).

The highest TSL elevations showed a temporal variability in the different years 8 (Table 3). The highest TSL elevations were found
in 2019 for Bors and 2018 for all other glaciers (Table 3). Lowest elevations were reached in 2019 at Pizzo Ferrè and Gran Zebrù,
2020 at Seewjinen and Solda Orientale and 2021 at Bors (Fig. 4f). The temporal evolution of the mapped end of summer snow lines
by the year is different for each glacier, only Seewjinen and Pizzo Ferrè have about similar trends (Fig. 4f) despite their different loca-
tion (Fig. 1). The maximum variability in a single year occurred at Solda Orientale glacier, with a difference between the highest and
lowest (respectively in 2018 and 2020) of 232 m (Fig. 4e and f). Differently, at the other glaciers the interannual variability was al-
ways ≤100 m.
It must be considered that what we calculated is a mean value of the elevation of the TSL, but the elevation variability of the snow-
line on an individual glacier can reach several 100m, also visible in Fig. 4.

5
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Fig. 3. Percentage area decrease (%) of the five glaciers between 2017 and 2021.

Table 2
Glacier area (km2) of the five glaciers from 2017 to 2021, total area decrease (km2), percentage area decrease (%) and mean change rate (%/year).

Area (km2) Bors Seewjinen Pizzo Ferrè Gran Zebrù Solda Or.

2017 0.738 ± 0.019 1.349 ± 0.030 0.445 ± 0.018 0.667 ± 0.018 0.957 ± 0.025
2018 0.715 ± 0.019 1.321 ± 0.032 0.427 ± 0.018 0.662 ± 0.018 0.940 ± 0.025
2019 0.700 ± 0.019 1.307 ± 0.032 0.410 ± 0.020 0.653 ± 0.018 0.930 ± 0.025
2020 0.696 ± 0.019 1.300 ± 0.032 0.395 ± 0.019 0.645 ± 0.018 0.918 ± 0.026
2021 0.681 ± 0.019 1.281 ± 0.032 0.385 ± 0.019 0.641 ± 0.018 0.910 ± 0.026
Total area decrease (km2) 0.056 ± 0.019 0.068 ± 0.032 0.061 ± 0.019 0.026 ± 0.018 0.047 ± 0.026
Percentage area decrease (%) 7.6 ± 2.6 5.0 ± 2.4 13.6 ± 4.3 3.9 ± 2.7 4.9 ± 2.7
Mean change rate (%/year) 1.91 1.26 3.40 0.98 1.22

4.3. Image classification

Here we report the image classification results of the two glaciers (Gran Zebrù and Seewjinen glaciers). Although PlanetScope im-
ages are daily, a limited number of images are useable, as images with cloud cover or other problems (as band saturation on snow ar-
eas) must be discarded. Only the images where ice/debris was exposed on the glaciers were classified, as in the previous and succes-
sive images the glaciers are 100% snow covered. A total number of 65 images were classifiable (not cloud covered and not affected by
other problems as pixel saturation) for Gran Zebrù glacier and 60 for Seewjinen glacier. At Gran Zebrù 205 images out of 390 days
(53%) were discarded due to cloud cover, while at Seewjinen 240 images out of 448 days (54%).
Ice/debris surfaces start being exposed earlier at Seewjinen glacier (between 04 June and 02 July in the different years) than at
Gran Zebrù glacier (between 26 June and 17 July). The ablation season usually ends at the end of September at both glaciers. Excep-
tions are 2017, when it ended earlier: on 29 August at Gran Zebrù and 7 September at Seewjinen and later at Gran Zebrù in 2018,
when it lasted until 26 October (Fig. 5). After snow depletion starts, snow cover decreases and ice and debris surfaces increase, with
different trends at the different glaciers and in the different years (Fig. 5). One or more snowfall events usually occur during the abla-
tion season at both glaciers. At Seewjinen glacier the snow covered area never decreased below 30%, while at Gran Zebrù it reached a
minimum of 7% in September 2018.
Comparing the snow covered area retrieved by the image classification and the SCAR obtained from the manually mapped TSLs on
the same images for Gran Zebrù and Seewjinen glaciers, we found a mean difference of ±6% of the total glacier area. The major dif-
ferences are found on September 04, 2019 at Gran Zebrù glacier, where the SCAR is higher than the snow covered area retrieved by
image classification, due to the fact that above the snow line some debris covered patches are also present (Fig. 1 Supplementary),
and on September 28, 2018 at Seewjinen glacier, with the snow covered area retrieved by image classification being higher than the
SCAR by 11%, in this case it is the image classification that overestimates the snow cover area, due to the incorrect classification of
part of the glacier ice surface as snow.

5. Discussion
5.1. The use of PlanetScope satellite images for studying small glacier
The PlanetScope images offer advantages both in terms of spatial and temporal resolution in glacier investigations. Regarding
glacier mapping, we calculated small uncertainties, between 2.2 and 5.1% of the glacier area (Table 2). In our case annual glacier
changes are on average −1.5%, within the uncertainties of the mapped areas, it is therefore advisable to determine glacier area
changes with PlanetScope images with a temporal gap of at least 3 years so that the changes are greater than the uncertainties.
The high resolution of PlanetScope images improves the interpretation of the glaciers boundaries. Among the commonly used
satellite imagery, Sentinel-2 has the highest spatial resolution of 10 m, while Landsat 8 of 15 m (panchromatic) and 30 m (Visible
(VIS)/NIR/SWIR), ten times the pixel size of PlanetScope.

6
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Fig. 4. Highest position of the transient snowline at the end of the ablation season for each year between 2017 and 2021 for: a) Bors glacier; b) Seewjinen glacier; c)
Pizzo Ferrè glacier; d) Gran Zebrù glacier; e) Solda Orientale glacier; f) Box plot showing the interannual variability of the highest TSL elevation for the five glaciers
between 2017 and 2021 (for each boxplot, the central mark is the mean value, the edges of the box are the means ± one standard deviation and the whiskers extend
to maximum and minimum values). The images in the background are the PlanetScope images from 2017 used to map the respective snowlines: a) 24/08/2017; b)
24/08/2017; c) 29/08/2017; d) 29/08/2017; e) 30/08/2017.

In Fig. 6, as an example, it is possible to see the differences in spatial detail of a PlanetScope image (3 m resolution, Fig. 6a and d),
a Sentinel-2 (Fig. 6b and e) and a Landsat 8 image (Fig. 6c and f) of the same glaciers, Gran Zebrù glacier and Solda Orientale glacier,
on the same day (27/08/2019), although Sentinel-2 compared to Landsat has already a much higher resolution.
The advantages of high resolution images for small gaciers have been already demonstrated in the literature, for example Fischer
et al. (2014) showed that using 30 m resolution Landsat images can lead to a misclassification of more than 25% of the area of very
small glaciers (<0.5 km2). They found that using high-resolution images works in favor of accurate glacier mapping as they provide a
high level of visible detail and small-scale structures like crevasses, exposed glacier ice below snow or debris-cover, or abrupt changes
in the surface topography.

7
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Table 3
For each glacier are reported: date of the PlanetScope image where the snowline was drawn, highest TSL elevation (m a.s.l.) from 2017 to 2021, TSL anomaly re-
spect to the median elevation of the glacier (m) and snow cover area ratio (SCAR) (%). Also the mean value for each glacier is reported.

Year Date Elev. (m a.s.l.) Anom. (m) SCAR (%)

Bors 2017 24-Aug 3463 52 25

2018 27-Sep 3465 54 26
2019 30-Sep 3472 61 22
2020 15-Sep 3440 29 33
2021 09-Sep 3372 −39 62
Mean 3442 31 34
Seewjinen 2017 24-Aug 3047 36 32
2018 28-Sep 3077 66 24
2019 03-Sep 3035 24 37
2020 15-Sep 2994 −17 44
2021 18-Sep 3049 38 32
Mean 3040 29 34
Pizzo Ferrè 2017 29-Aug 2980 149 4
2018 28-Sep 2990 159 3
2019 04-Sep 2903 72 14
2020 17-Sep 2925 94 18
2021 18-Sep 2981 150 3
Mean 2956 125 8
Gran Zebrù 2017 29-Aug NA NA 0
2018 30-Sep NA NA 0
2019 04-Sep 3216 26 31
2020 18-Sep 3258 68 24
2021 18-Sep 3233 43 23
Mean 3236 46 16
Solda Orientale 2017 30-Aug 3199 214 19
2018 21-Aug 3211 226 16
2019 03-Sep 3105 120 36
2020 19-Sep 2979 −6 51
2021 18-Sep 2992 7 51
Mean 3097 112 35

Moreover using higher resolution images for glacier mapping also enhances the visibility of debris-covered glacier parts and per-
mits to obtain more accurate outlines (Paul et al., 2017).
However, manual mapping of glacier outlines and snowlines introduces some subjectivity and depends on the analyst performing
it. In particular mapping debris-covered areas is challenging and can lead to strong differences in interpretation of the outlines (e.g.
Paul et al., 2020). Mölg et al. (2018) performed a multiple digitizing experiment, finding that if a glacier is affected by both shadow
and debris cover and is also small, the interpretation of different analysts can differ by up to 70%. Errors related to snowlines digitized
manually on satellite images depend not only on their spatial resolution and DEMs accuracy, but mainly on misinterpretations of the
analyst, in particular in case of small-scale snow cover variability.
The higher spatial resolution compared to other satellites is also useful for a more accurate interpretation of the surface facies.
Other than the high spatial resolution, the main advantage of using PlanetScope images is the daily revisit time. The short revisit
time increases the probability of obtaining data also in regions that are frequently cloud-covered. Indeed, when mapping glaciers out-
lines or the TSL, cloud-free images from the end of the ablation season and with minimum snow cover are needed, but when using
other satellites with lower temporal resolution it is often not possible to obtain images from a specific period with these conditions
(Paul et al., 2020). As the revisit time of Sentinel-2 is 5 days, and it's 8 days for Landsat (combining Landsat 8 and Landsat 9), the
amount of available data is lower than for PlanetScope. Moreover, thanks to their daily availability, PlanetScope images can be used
to monitor the glacier surface at sub-seasonal time scales. For example at Gran Zebrù, out of 390 days of the abation seasons
2017–2021, 65 PlanetScope images were useable (17%), while the number of Sentinel-2 and Landsat 8 images were respectively 40
(10%) and 10 (3%). At Seewjinen out of 448 days, 60 PlanetScope images were useable (13%), 30 Sentinel-2 (7%) and 9 Landsat 8
(2%). This higher availability of data can be used to follow closely the evolution of snow cover extent, as we show in Fig. 5.
Web-map services such as the World Imagery layer of the ESRI Basemap can offer images of even higher resolution than the Plan-
etScope images, but can't offer their near-daily temporal availability.
Although Planet's satellites are commercial, a limited and non-commercial access to data is offered through the Education and Re-
search Program for research purposes. A download quota of 5000 km2/month is offered, not allowing large scale studies. Therefore,
PlanetScope images can be used for the detailed study of a limited number of glaciers rather than mapping thousands of glaciers for
an inventory.
PlanetScope images have also some other disadvantages. One is the absence of a SWIR band, not permitting the automated map-
ping of glaciers using band ratio methods. Another problem is pixels saturation occurring principally on snow areas, already reported
by Ghuffar (2018) and Aati & Avouac (2020). But the NIR band is usually not affected by this problem and could be used alone for

8
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Fig. 5. Evolution of the different surface types (snow, ice, debris) during the 2017–2021 ablation seasons at Gran Zebrù (a) and Seewjinen (b) glaciers.

9
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Fig. 6. Satellite images of Gran Zebrù glacier (a, b, c) and Solda Orientale glacier (d, e, f) acquired on 27/08/2019: a and d) PlanetScope image; b and e) Sentinel-2 im-
age; c and f) Landsat 8 image. Outlines are the 2019 outlines mapped on PlanetScope images.

some applications. In addition, the first PlanetScope satellites were launched in 2016 and so long timeseries of data are not yet avail-
able.

5.2. Comparison with previous glacier outlines

The glacier outlines obtained from PlanetScope images can be compared to other available glacier outlines of the same glaciers.
The most recent glacier outlines found are from the glacier inventory of the Alps, published by Paul et al. (2019), where glacier out-
lines were mapped on Sentinel-2 data (10 m resolution) acquired in 2015, but also in 2016 and 2017 in case they were not available
due to cloud cover (Fig. 7). They performed automated mapping of clean glacier ice using the band ratio method, with manual correc-
tions for areas covered by debris, clouds, shadow regions and wrongly mapped water bodies (Paul et al., 2020). For the five glaciers
considered in this study, only one outline is from 2015 (Seewjinen glacier), while three are from 2016 (Pizzo Ferrè, Gran Zebrù and
Solda Orientale glaciers) and one from 2017 (Bors glacier). The outlines were not compared with the ones published in the Randolph
Glacier Inventory (Version 6.0), a global inventory of glacier outlines, as they refer to 2003 and are therefore not comparable.
In Table 4 the glacier areas from 2017 from PlanetScope and the glacier areas from the glacier inventory of the Alps are reported
for the five glaciers. All the glacier areas reported by Paul et al. (2019) are smaller than the ones reported in this study, although they
refer to the same year (Bors glacier) or to the two previous years (the other four glaciers) (Table 4, Fig. 7). In particular, the area dif-
ference is small for Seewjinen and Pizzo Ferrè glaciers (only 1% smaller), while it increases for Gran Zebrù, Bors and Solda Orientale
glaciers (13% smaller) (Table 4).
Observing the three glaciers with the greater area difference, Bors, Gran Zebrù and Solda Orientale (Fig. 7a, d and 7e), it is possi-
ble to see that some areas that we mapped as debris-covered ice were not mapped as glacier areas by Paul et al. (2019), suggesting
that the higher resolution of PlanetScope images helps identify the debris-covered areas. Moreover, in our case, the inclusion and de-
lineation of the debris covered areas at the eastern tongue of Gran Zebrù glacier and at Solda Orientale glacier not mapped by Paul et
al. (2019) was favoured by the analyst's knowledge of the sites.

5.3. Glacier evolution and comparison with other studies on alpine glaciers
Some studies carried out in the Alps showed that small glaciers are showing largest relative area decrease respect to large glaciers.
For example, for the Italian glaciers Smiraglia et al. (2015) reported strongest area reduction for small glaciers (<1 km2) in the period
from 1959 to 1961 to 2005–2011. Also, Carturan et al. (2013a) found a tendence to a greater relative area loss with decreasing initial

10
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Fig. 7. Glacier outlines mapped on 2017 PlanetScope images (yellow) and glacier outlines from the glacier inventory of the Alps (a) 2017; b)2015; c,d,e) 2016) by
Paul et al. (2019) (purple) of the five glaciers: a) Bors glacier; b) Seewjinen glacier; c) Pizzo Ferrè glacier; d) Gran Zebrù glacier; e) Solda Orientale glacier. The images
in the background are PlanetScope images from the end (23–30) of August 2017. (For interpretation of the references to colour in this figure legend, the reader is re-
ferred to the Web version of this article.)

Table 4
Glacier areas (km2) of the five investigated glaciers mapped on 2017 PlanetScope images and the values published in the glacier inventory of the Alps by Paul et al.
(2019) mapped on 2015/2016/2017 Sentinel-2 images.

Bors Seewjinen Pizzo Ferrè Gran Zebrù Solda Orientale

Area 2017 Pl. (km2) 0.738 1.349 0.445 0.667 0.957

Area Alps Inventory (km2) 0.645 1.339 0.442 0.579 0.829
Area difference (%) −12.6 −0.7 −0.7 −13.2 −13.4

11
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

size for the glaciers of the Ortles-Cevedale (Italian Alps) in the period 1987–2009. In the Swiss Alps, Linsbauer et al. (2021), found an
area decrease of −60% for glaciers <0.1 km2, -39% for glaciers 0.1–1 km2, -21% for glaciers 1–10 km2 and -13% for glaciers >
10 km2 in the period between 1973 and 2016. Differently, the resilience of some very small glaciers was reported in some areas of the
Alps, as in the Julian Alps (Eastern Italian Alps) (by Carturan et al., 2013b; Colucci, 2016; Colucci et al., 2021 and De Marco et al.,
2020) and in the Orobie (Central Italian Alps) (by Scotti et al., 2014), zones characterized by high mean annual precipitation.
The five glaciers considered in this study showed different rates of area loss. The glaciers have different characteristics, as different
area, slope, hypsometry, aspect, quantity of supraglacial debris, glacier thickness and surrounding topography, and are located in ar-
eas with different local meteorological conditions. All these factors surely concur in driving their evolution, but with only five glaciers
it is not possible to identify the factors driving it.
In our sample, Pizzo Ferrè glacier is the one that lost the larger relative area (−14%) and is both the smallest, covering an area of
0.445 km2 in 2017 and thus being considered a very small glacier (<0.5 km2), and the one having the lowest median elevation
(2831 m a.s.l.).
The total measured area changes of the five glaciers, showed an area loss of 6.2% of the initial area between 2017 and 2021, with a
rate of decrease of about −1.5% per year, are in agreement with other studies carried out in the alpine region, although they refer to
previous periods respect to the one investigated in this study. For example, D'Agata et al. (2018) reported a mean area change of
−1.6% per year between 1991 and 2007 for the glaciers of the Sondrio Province (Italy) (to which Pizzo Ferrè glacier and Gran Zebrù
glacier belong). Knoll and Kerschner (2009) found an area change of −1.4% per year in South Tirol (Italy) and in particular of −1.1%
in the Ortles Cevedale group (to which Solda Orientale glacier belong) in the period 1983–2006. Huss et al. (2012) found an area de-
crease of −1.7% from 1999 to 2008 for 36 glaciers of the Swiss Alps. Paul et al. (2020) reported an area loss of −15% (−1.3% per
year) from 2003 to 2015/2016 for glaciers in the Alps, while Sommer et al. (2020) about −1.8% per year between 2000 and 2014.
Considering that these studies refer to glaciers of every size and not only to small glaciers, we did not find in this study a greater area
loss rate for small glaciers respect to larger ones.
The temporal variability of the highest SLA doesn't follow the same temporal trends of summer temperatures or precipitation dur-
ing the ablation season, suggesting that other factors other than precipitation and temperature might contribute to determine it, for
example wind-drifting or avalanching of snow, that influence accumulation on glaciers (Kuhn, 1995) or the cloud cover, that can in-
fluence ablation.
Extending the length of the timeseries, using remote sensing images of satellites operative before 2017, increasing the number of
glaciers investigated but also using a different kind of analysis, as distributed mass balance models (considering shading, debris,
avalanches, etc.) could improve the understanding of the topographic and climatic factors that influence the evolution of small glaci-
ers.

6. Conclusions
In this study we showed that PlanetScope satellite images, thanks to their high temporal (daily) and spatial (3 m) resolution are
valuable for studying small glaciers, offering different advantages compared to sensors such as Sentinel-2 or Landsat 8/9, despite
some disadvantages (missing of the SWIR band, small scene size and limited download quota). We were able to retrieve the glacier
area of five small glaciers in the Alps with low uncertainties (2.2–5.1% of the area) and to see their evolution from 2017 to 2021. They
lost between 4 and 14% of the area, with the smallest glacier being the one that lost more area. The highest TSLs showed interannual
variability, being probably influenced both by temperatures and precipitation, with different temporal trends at the different glaciers.
Monitoring the evolution of the glacier surface of two of the glaciers during the ablation season, we saw differences in the length of
the ablation season and in the evolution of the snow cover. PlanetScope images, with their daily availability offer an increased num-
ber of images compared to other commonly used satellites, allowing to study glaciers not only annually but also sub-seasonally.

Author statement
The authors Mauro Guglielmin (MG), Martin Hoelzle (MH) and Giulia Tarca (GT) are responsible respectively of: Conceptualiza-
tion, Methodology, Validation, (All), Writing review and Editing (MG and GT), Formal analyses, Investigation, Data Curation, Writing
original draft and Visualization (GT), Supervision, Funding acquisition and Project Administration (MG).

Ethics statement
I and the other authors agreed on the ethics policy of the Journal.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to
influence the work reported in this paper.

Data availability
Data will be made available on request.

12
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Acknowledgements
We are thankful to Planet Labs for providing access to the satellite images used in this study through the Education and Research
program. We also thank ARPA Piemonte, ARPA Lombardia and Provincia autonoma di Bolzano - Alto Adige for providing meteoro-
logical data and Geoportale della Lombardia, Geoportale Alto Adige, Geoportale Piemonte and Federal Office of Topography swis-
stopo for providing the digital elevation models.
We also thank the editor and the two anonymous reviewers who allowed us to significantly improve the paper with their sugges-
tions.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.rsase.2023.101013.

References
Aati, S., Avouac, J.P., 2020. Optimization of optical image geometric modeling, application to topography extraction and topographic change measurements using
PlanetScope and SkySat imagery. Rem. Sens. 12 (20), 3418.
Auer, I., Böhm, R., Jurkovic, A., Lipa, W., Orlik, A., Potzmann, R., Schöner, W., Ungersbock, M., Matulla, C., Briffa, K., Jones, P.D., Efthymiadis, D., Brunetti, M., Nanni,
T., Maugeri, M., Mercalli, L., Mestre, O., Moisselin, J.-M., Begert, M., Müller-Westermeier, G., Kveton, V., Bochnicek, O., Statsny, P., Lapin, M., Szalai, S.,
Szentimrey, T., Cegnar, T., Dolinar, M., Gajic-Capka, M., Zaninovic, K., Majstorovic, Z., Nieplova, E., 2007. HISTALP—historical instrumental climatological
surface time series of the Greater Alpine Region. Int. J. Climatol.: J. R. Meteorol. Soc. 27 (1), 17–46.
Azzoni, R.S., Fugazza, D., Zerboni, A., Senese, A., D’Agata, C., Maragno, D., Carzaniga, A., Cernuschi, M., Diolaiuti, G.A., 2018. Evaluating high-resolution remote
sensing data for reconstructing the recent evolution of supra glacial debris: a study in the Central Alps (Stelvio Park, Italy). Prog. Phys. Geogr. Earth Environ. 42 (1),
3–23.
Barandun, M., Huss, M., Sold, L., Farinotti, D., Azisov, E., Salzmann, N., Usubaliev, R., Merkushkin, A., Hoelzle, M., 2015. Re-analysis of seasonal mass balance at
Abramov glacier 1968–2014. J. Glaciol. 61 (230), 1103–1117.
Carturan, L., Filippi, R., Seppi, R., Gabrielli, P., Notarnicola, C., Bertoldi, L., Paul, F., Rastner, P., Cazorzi, F., Dinale, R., Dalla Fontana, G., 2013a. Area and volume loss
of the glaciers in the Ortles-Cevedale group (Eastern Italian Alps): controls and imbalance of the remaining glaciers. Cryosphere 7 (5), 1339–1359.
Carturan, L., Baldassi, G.A., Bondesan, A., Calligaro, S., Carton, A., Cazorzi, F., Dalla fontana, G., Francese, R., Guarnieri, A., Milan, N., Moro, D., Tarolli, P., 2013b.
Current behaviour and dynamics of the lowermost Italian glacier (Montasio Occidentale, Julian Alps). Geogr. Ann. Phys. Geogr. 95 (1), 79–96.
Chand, P., Jain, S.K., Thakur, H.P., Kumar, S., Sharma, M.C., 2020. Recessional pattern and surface elevation change of the Parvati Glacier, North-Western Himalaya
(1965-2018) using remote sensing. Int. J. Rem. Sens. 41 (24), 9360–9392.
Colucci, R.R., 2016. Geomorphic influence on small glacier response to post-Little Ice Age climate warming: Julian Alps, Europe. Earth Surf. Process. Landforms 41 (9),
1227–1240.
Colucci, R.R., Žebre, M., Torma, C.Z., Glasser, N.F., Maset, E., Del Gobbo, C., Pillon, S., 2021. Recent increases in winter snowfall provide resilience to very small
glaciers in the Julian Alps, Europe. Atmosphere 12 (2), 263.
D’Agata, C., Bocchiola, D., Soncini, A., Maragno, D., Smiraglia, C., Diolaiuti, G.A., 2018. Recent area and volume loss of Alpine glaciers in the Adda River of Italy and
their contribution to hydropower production. Cold Reg. Sci. Technol. 148, 172–184.
De Marco, J., Carturan, L., Piermattei, L., Cucchiaro, S., Moro, D., Dalla Fontana, G., Cazorzi, F., 2020. Minor imbalance of the lowermost Italian glacier from 2006 to
2019. Water 12 (9), 2503.
Fischer, M., Huss, M., Barboux, C., Hoelzle, M., 2014. The new Swiss Glacier Inventory SGI2010: relevance of using high-resolution source data in areas dominated by
very small glaciers. Arctic Antarct. Alpine Res. 46 (4), 933–945.
Frei, C., Schär, C., 1998. A precipitation climatology of the Alps from high-resolution rain-gauge observations. Int. J. Climatol.: J. R. Meteorol. Soc. 18 (8), 873–900.
Ghuffar, S., 2018. DEM generation from multi satellite PlanetScope imagery. Rem. Sens. 10 (9), 1462.
Huss, M., Hock, R., Bauder, A., Funk, M., 2012. Conventional versus reference-surface mass balance. J. Glaciol. 58 (208), 278–286.
Huss, M., Sold, L., Hoelzle, M., Stokvis, M., Salzmann, N., Farinotti, D., Zemp, M., 2013. Towards remote monitoring of sub-seasonal glacier mass balance. Ann. Glaciol.
54 (63), 75–83.
Huss, M., Fischer, M., 2016. Sensitivity of very small glaciers in the Swiss Alps to future climate change. Front. Earth Sci. 4, 34.
Isotta, F.A., Frei, C., Weilguni, V., Perčec Tadić, M., Lassegues, P., Rudolf, B., Pavan, V., Cacciamani, C., Antolini, G., Ratto, S.M., Maraldo, L., Micheletti, S., Bonati, V.,
Lussana, C., Ronchi, C., Panettieri, C.E., Marigo, G., Vertačnik, G., 2014. The climate of daily precipitation in the Alps: development and analysis of a high-
resolution grid dataset from pan-Alpine rain-gauge data. Int. J. Climatol. 34 (5), 1657–1675.
Kääb, A., Leinss, S., Gilbert, A., Bühler, Y., Gascoin, S., Evans, S.G., Bartelt, P., Berthier, E., Brun, F., Chao, W.-A., Farinotti, D., Gimbert, F., Guo, W., Huggel, C., Kargel,
J.S., Leonard, G.J., Tian, L., Treichler, D., Yao, T., 2018. Massive collapse of two glaciers in western Tibet in 2016 after surge-like instability. Nat. Geosci. 11 (2),
114–120.
Knoll, C., Kerschner, H., 2009. A glacier inventory for South Tyrol, Italy, based on airborne laser-scanner data. Ann. Glaciol. 50 (53), 46–52.
Kuhn, M., 1995. The mass balance of very small glaciers. Z. Gletscherkd. Glazialgeol. 31 (1), 171–179.
Linsbauer, A., Huss, M., Hodel, E., Bauder, A., Fischer, M., Weidmann, Y., Bärtschi, H., Schmassmann, E., 2021. The new Swiss Glacier Inventory SGI2016: from a
topographical to a glaciological dataset. Front. Earth Sci. 9.
Millan, R., Mouginot, J., Rabatel, A., Jeong, S., Cusicanqui, D., Derkacheva, A., Chekki, M., 2019. Mapping surface flow velocity of glaciers at regional scale using a
multiple sensors approach. Rem. Sens. 11 (21), 2498.
Mölg, N., Bolch, T., Rastner, P., Strozzi, T., Paul, F., 2018. A consistent glacier inventory for Karakoram and Pamir derived from Landsat data: distribution of debris
cover and mapping challenges. Earth Syst. Sci. Data 10 (4), 1807–1827.
Paul, F., Barrand, N.E., Baumann, S., Berthier, E., Bolch, T., Casey, K., Frey, H., Joshi, S.P., Konovalov, V., Le Bris, R., Mölg, N., Nosenko, G., Nuth, C., Pope, A.,
Racoviteanu, A., Rastner, P., Raup, B., Scharrer, K., Steffen, S., Winsvold, S., 2013. On the accuracy of glacier outlines derived from remote-sensing data. Ann.
Glaciol. 54 (63), 171–182.
Paul, F., Bolch, T., Briggs, K., Kääb, A., McMillan, M., McNabb, R., Nagler, T., Nuth, C., Rastner, P., Strozzi, T., Wuite, J., 2017. Error sources and guidelines for quality
assessment of glacier area, elevation change, and velocity products derived from satellite data in the Glaciers_cci project. Remote Sens. Environ. 203, 256–275.
Paul, F., Rastner, P., Azzoni, R.S., Diolaiuti, G., Fugazza, D., Le Bris, R., Nemec, J., Rabatel, A., Ramusovic, M., Schwaizer, G., Smiraglia, C., 2019. Glacier inventory of
the Alps from Sentinel-2, shape files. PANGAEA. https://doi.org/10.1594/PANGAEA.909133. Supplement to: Paul, et al. (2020). Glacier shrinkage in the Alps
continues unabated as revealed by a new glacier inventory from Sentinel-2. Earth System Science Data, 12(3), 1805-1821.
Paul, F., Rastner, P., Azzoni, R.S., Diolaiuti, G., Fugazza, D., Le Bris, R., Nemec, J., Rabatel, A., Ramusovic, M., Schwaizer, G., Smiraglia, C., 2020. Glacier shrinkage in
the Alps continues unabated as revealed by a new glacier inventory from Sentinel-2. Earth Syst. Sci. Data 12 (3), 1805–1821.
Planet, 2021. Planetscope Product Specifications. https://assets.planet.com/docs/Planet_Combined_Imagery_Product_Specs_letter_screen.pdf.
Pudełko, R., Angiel, P.J., Potocki, M., Jędrejek, A., Kozak, M., 2018. Fluctuation of glacial retreat rates in the eastern part of Warszawa Icefield, king George Island,
Antarctica, 1979–2018. Rem. Sens. 10 (6), 892.
Qayyum, N., Ghuffar, S., Ahmad, H.M., Yousaf, A., Shahid, I., 2020. Glacial lakes mapping using multi satellite PlanetScope imagery and deep learning. ISPRS Int. J.
Geo-Inf. 9 (10), 560.

13
G. Tarca et al. Remote Sensing Applications: Society and Environment 32 (2023) 101013

Racoviteanu, A.E., Paul, F., Raup, B., Khalsa, S.J.S., Armstrong, R., 2009. Challenges and recommendations in mapping of glacier parameters from space: results of the
2008 Global Land Ice Measurements from Space (GLIMS) workshop, Boulder, Colorado, USA. Ann. Glaciol. 50 (53), 53–69.
Racoviteanu, A.E., Rittger, K., Armstrong, R., 2019. An automated approach for estimating snowline altitudes in the Karakoram and eastern Himalaya from remote
sensing. Front. Earth Sci. 7, 220.
Ragettli, S., Bolch, T., Pellicciotti, F., 2016. Heterogeneous glacier thinning patterns over the last 40 years in Langtang Himal, Nepal. Cryosphere 10 (5), 2075–2097.
Scotti, R., Brardinoni, F., Crosta, G.B., 2014. Post-LIA glacier changes along a latitudinal transect in the Central Italian Alps. Cryosphere 8 (6), 2235–2252.
Shaw, T.E., Ulloa, G., Farías-Barahona, D., Fernandez, R., Lattus, J.M., McPhee, J., 2021. Glacier albedo reduction and drought effects in the extratropical Andes,
1986–2020. J. Glaciol. 67 (261), 158–169.
Smiraglia, C., Azzoni, R.S., D’Agata, C., Maragno, D., Fugazza, D., Diolaiuti, G.A., 2015. The evolution of the Italian glaciers from the previous data base to the New
Italian Inventory. Preliminary considerations and results. Geogr. Fis. Din. Quaternaria 38 (1), 79–87.
Sommer, C., Malz, P., Seehaus, T.C., Lippl, S., Zemp, M., Braun, M.H., 2020. Rapid glacier retreat and downwasting throughout the European Alps in the early 21 st
century. Nat. Commun. 11 (1), 1–10.
Steiner, J.F., Kraaijenbrink, P.D., Jiduc, S.G., Immerzeel, W.W., 2018. Brief communication: the Khurdopin glacier surge revisited–extreme flow velocities and
formation of a dammed lake in 2017. Cryosphere 12 (1), 95–101.
Tarca, G., Guglielmin, M., 2022a. Using ground-based thermography to analyse surface temperature distribution and estimate debris thickness on Gran Zebrù glacier
(Ortles-Cevedale, Italy). Cold Reg. Sci. Technol. 103487.
Tarca, G., Guglielmin, M., 2022b. Evolution of the sparse debris cover during the ablation season at two small Alpine glaciers (Gran Zebrù and Sforzellina, Ortles-
Cevedale group). Geomorphology 409, 108268.
Zemp, M., 2006. Glaciers and Climate Change: Spatio-Temporal Analysis of Glacier Fluctuations in the European Alps after 1850. Doctoral dissertation, University of
Zurich.

14