Geomorphology of the Maladeta massif (Central Pyrenees): the traces of the last remaining glaciers

ABSTRACT

We present a 1:15,000 geomorphological map of the Maladeta massif in the Central Pyrenees. Our methodology includes fieldwork, the analysis of aerial photographs from 1956 to 2015, and the use of drone flights from 2020 to 2023. The study area consists mainly of granodiorite but also has outcrops of limestone, shales, and quartzites at lower elevations. The landscape of the Maladeta massif is primarily shaped by glacial and periglacial processes, with fluvial, karstic, and hillslope dynamics manifesting in the lower regions. The assessment of moraine thickness has facilitated the determination of maximum glacier thickness during the Little Ice Age (LIA) and has shown that there is no correlation between ice thickness and glacier extent. After the LIA, the massif experienced a continuous glacier retreat. Currently, only 21.4% of the glacier area observed in 1956 remains.

KEYWORDS:

• Geomorphology
• Maladeta massif
• Pyrenees
• glaciers
• moraines
• Little Ice Age

1. Introduction

The Ésera valley, where the Maladeta massif is located, has been the subject of numerous geomorphological studies. Bordonau (1992, 1993) studied the glacial geomorphology of the southern slope of the Maladeta massif, while Martínez de Pisón (1989, 1990) described the geomorphology of the headwaters of the Ésera valley. The geomorphological map of Benasque was addressed by García-Ruiz et al. (1992), and Chueca-Cía and Julián-Andrés (2008) studied the geomorphology of Alta Ribagorza. All of them mapped the Maladeta massif at a resolution of 1:50,000. González-García (2013) mapped the periglacial processes within the Maladeta cirque at a resolution of 1:10,000. In particular, Crest et al. (2017) and Reixach (2022) carried out various studies on the dating of moraines, till and some polished bedrocks of the Maladeta massif using OSL methods, which provided insights into the glaciers during the Little Ice Age (LIA), Younger Dryas and Oldest Dryas periods, although they only focused on moraine sites.

The aim of this study is to produce a detailed geomorphological map (1:15,000) of the Maladeta massif, focusing mainly on the landforms modelled by glacial activity and retreat since the last glacial cycle, although other landforms such as periglacial, karstic, fluvial, hillslope, structural and anthropogenic landforms are described. The geomorphological mapping in this mountain area makes it possible to understand the sequence of geomorphological processes and the evolution of climate. Special attention is given to the detailed assessment of the LIA moraines and the evolution of ice cover since 1956. This map provides comprehensive details and serves as a basis for various future studies that include landscape evolution, paleoclimate reconstruction, landform chronology, and assessment of plant and animal distribution and ecosystem services.

2. Study area

The Maladeta massif is located in the Central Pyrenees (Figure 1) and is one of the highest massifs in this mountain range with more than 40 peaks over 3000 m above sea level (a.s.l.), including the summit of Aneto (3404 m a.s.l.), the highest point in the Pyrenees. Geologically, the massif is located in the Axial Zone of the Pyrenees. This consist of (1) granites that were formed during the Variscan orogeny and predominate in the upper regions of the massif. And (2) the Devonian and Carboniferous limestones, shales, and quartzites, which were strongly influenced by the Variscan and Alpine orogeny’s and predominate in the lower areas (García-Sansegundo et al., 2013; Ríos-Aragüés et al., 2002) (Figure 1c).

Figure 1. Location map of the Maladeta massif. (a) Location of the Pyrenees in Europe (Google Earth map). (b) Maladeta massif located in the Pyrenees; the pin indicates the location of the Maladeta massif in the Pyrenees. (c) Geological map of the Maladeta massif (modified data from the Instituto Geológico y Minero de España, IGME).

(a) A pink square mark the Pyrenees in the European context. (b) The surface in the square is expanded to located Maladeta massif with a pin. (c) This map shows the geology of the Maladeta massif, the main peaks, ravines, edges, thrust, faults and glaciers are represented. The four lithology of the massif (Devonian shales and limestones, carboniferous shales and quartzite, carboniferous limestone and granodiorite are difference by different symbology.

The landscape of the Maladeta massif is primarily shaped by glacial and periglacial processes, especially in granites, while limestones often show karstic landforms. Currently, the 0°C isotherm for this massif around 3200 m a.s.l. (considering the AEMET database of Besurta and Renclusa weather station, Clima y Nieve database from Llanos del Hospital station and Posets-Maladeta Natural Park database from the Aneto station, and a lapse rate of 0.525°C every 100 m). The mean annual precipitation is over 1300 mm (measured at the Renclusa station, 2140 m a.s.l., for the period 2008–2022; AEMET). The environmental conditions, altitude and orientation of the massif have contributed to the preservation of some of the last remaining glaciers and permafrost in the Pyrenees (Izagirre et al., under review; Rico, 2019; Serrano et al., 2019; Vidaller et al., 2021).

3. Methods

3.1. Mapping methods

The production of the geomorphological map for the Maladeta massif was based on a multi-layered approach:

1. The initial delineation of geomorphological landforms (geomorphological landforms are listed in Table S1 in the Supplementary Material) was based on photograms from the Centro Nacional de Información Geográfica de España (CNIG) from 1990, which were examined using a TOPCON stereoscope.

2. This preliminary map was digitised with ArcGIS software, using a digital elevation model (DEM) with a spatial resolution of 2 m obtained from the CNIG and a hillshade generated in ArcGIS based on the same DEM. In addition, orthophotos from several years (1956, 1973, 2005, 2006, 2007, 2008, 2009, 2011, 2012 and 2015), also obtained from the CNIG, were integrated.

3. The landforms identified in the map were compared with previous lower resolution maps for the same area (Chueca-Cía & Julián-Andrés, 2008; García-Ruiz et al., 1992; González-García et al., 2011; Martínez de Pisón, 1989, 1990).

4. The initial map was thoroughly refined during fieldwork campaigns (from 2020 to 2023) that covered a large part of the mapped area. High-resolution (centimetric) aerial images from drones were used for some remote and inaccessible regions.

5. The collected geomorphological information was then digitally integrated into the final map, adopting the symbology of Martín-Serrano et al. (2004).

At the same time, the geological map of IGME 1:50,000 (MAGNA 50; sheets 148 Vielha and 180 Benasque) was downloaded and subsequently corrected during several fieldwork campaigns (Figure 1c). The final representation of the study, referred to as the Main Map, is a composite comprising a hillshade with a spatial resolution of 5 m, in which the geomorphological landforms are superimposed, together with other insets related to the geographical location and the geological map and interpretation.

3.2. LIA moraine thickness

A moraine thickness map was created to illustrate the maximum glacial thickness of each glacier during the LIA, focusing on well-defined moraines, including net boundaries and preserved ridges. The process was carried out using ArcGIS software and involved the following steps:

1. The polygons that delineated the moraines were converted to lines using the Polygon-To-Line tool in ArcGIS.

2. These lines were then converted to a raster format with a resolution of 1 m/pixel using the Feature-To-Raster function.

3. Subsequently, each pixel within these raster’s was replaced by points using the Raster-To-Points function.

4. To simplify the calculations, a selection procedure was implemented in which one out of every 20 points was selected for inclusion in the final dataset.

An analogous methodology was used for the moraine ridges, omitting the first step. The Extract-Multi-Value-To-Points tool facilitated the determination of elevation for each point using the 2 m DEM. Moraine thickness was determined by calculating the elevation difference between a point of the limit of the moraine and the nearest point on the ridge. These data points were interpolated to create an individual moraine delimitation and construct the raster map.

3.3. Glacier evolution since 1956

The shrinkage of glaciers after the LIA has already been the subject of previous studies (Chueca et al., 2003, 2007; López-Moreno et al., 2006; Rico, 2019; Rico et al., 2017; Vidaller et al., 2021). Landsat satellite images, orthophotos from the CNIG’s Plan Nacional de Ortofotografía Aérea (PNOA), airborne LiDAR data from the Instituto Geográfico Nacional Español (IGN) and UAV surveys (Izagirre et al., under review; Vidaller et al., 2021, 2023), were used to reconstruct recent changes in glacier surfaces.

The glacier extent was manually delineated for the years 1956 (PNOA orthophoto), 2011 (LiDAR survey), 2020 and 2023 (both UAV surveys) (see methodology in Vidaller et al., 2021). In the case of the years 1990 and 2000, the glacier outlines were automatically defined by applying the Normalised Differential Snow Index (NDSI) with a ratio of 0.4 to Landsat imagery (Debnath et al., 2018). Details of the characteristics of each remote sensing source are in Table S2 in the Supplementary Material.

4. Results

4.1. Description of the geomorphological landforms

Considering the specific geographical and geological context, the geomorphological landforms mapped in this study (Main Map) are mainly characterised by glacial and periglacial processes. Furthermore, hillslope processes, although not very common, are also active in deglaciated areas.

4.1.1. General map overview

The Maladeta massif has been shaped by glacial processes since the Upper Pleistocene, with more than 96% of the massif showing glacial and periglacial landforms (Figure 2a), polished bedrocks and talus slopes dominate. The lithological composition determines the preservation and diversity of these geomorphological features. Granodiorite covers, approximately, 80% of the surface of the massif and favours the preservation of glacial landforms. In contrast, limestones are mainly found at lower elevations (Figure 1c), which favours the development of karst landforms and prevents the preservation of the glacial erosion landforms, due to the dissolution of carbonate, especially in cold climates.

Figure 2. Graph (a) shows the percentage of the surface of each landform groups in the Maladeta massif. Graph (b) shows the percentage of the surface covered by each glacial landform. Graph (c) shows the same variable but for the periglacial landform.

Three graphs show the surface that represent each feature. In the graph (a), each category corresponds to each geomorphological landform (glacial, periglacial, fluvial, karstic, hillslope and anthropogenic). Graph (b) shows the distribution of surface of each glacial landform (polished bedrock, till, lake, sheepback rock, peatbog, glacier, proglacial fan and ice-patch). Graph (c) shows the distribution of surface of each periglacial landform (debris slope, debris cone, rock glacier and protalus rampart).

4.1.2. Glacial landforms

In 2023, 52 ha of glacial ice are still present, distributed over three glaciers (Aneto, Maladeta East and Tempestades), which are located near the cirque walls and show clear signs of degradation (Vidaller et al., 2023). Glaciers with negligible movement or with an area of less than 2 ha are classified as ice-patches. Coronas became an ice-patch in the period 2011–2020 (Vidaller et al., 2021), while Barrancs, Maladeta West, and Secondary Aneto were declared ice-patches after the summer of 2023 (Izagirre et al., under review). Some of these ice-patches have experienced significant debris cover, accelerated by the rapid degradation of rock-wall permafrost (Rico et al., 2021).

Glacial features are particularly pronounced in the upper granitic zones. Polished bedrocks, which include smooth rock surfaces with strong jointing, erratic boulders, glacial groove marks, and subglacial calcites, cover more than 68% of this group (Figure 2b). Sometimes, these surfaces were eroded by former subglacial water flows, resulting in the formation of subglacial gorges (Figure 3e). Sheepback rocks, which are mainly found at lower elevations (<2800 m a.s.l.), occasionally resemble polished bedrocks when the feature is not well developed.

Figure 3. (a) Northern slope of the Maladeta massif, the valley is U-shaped (marked with continuous yellow lines) and has Alba and Forau Tancau hanging valleys (yellow dotted lines). (b) Aligned sinkholes outlined in red (filled with snow) in the Llausía valley, under the Tuca Blanca de Paderna. (c) Overdeepening basin of Llosas lake. (d) Proglacial fan (pink) and LIA moraines (purple) of the Maladeta glacier. (e) Subglacial gorge (marked in yellow) and polished bedrock near the Aneto glacier. (f) Rock glacier (pink) in the Cregüeña cirque.

(a) U-shaped landform of the Ésera valley and the hanging valley of Alba and Forau Tancau valleys are marked in yellow. (b) Sinkholes alienated in the bottom of the Llausía valley are marked in red. (c) Overdeepening basing of Llosas lake is marked in yellow. (d) LIA moraine of Maladeta glacier is marked in purple and the proglacial fan originated from this moraine is marked in pink. (e) A subglacial gorge over the polished bedrock in the Aneto cirque is marked in yellow. (f) The relict rock glacier of the Cregüeña cirque is marked in pink.

The second most common landforms are till deposits and moraines. The youngest moraines, associated with the LIA and located above 2500 m a.s.l. (depicted with darker colours on the Main Map), are the best preserved and thickest. On the northern side of the Maladeta massif, the LIA moraines have been affected by glacier melt run off, forming proglacial fans (Figure 3d).

All the lakes within the massif were formed by glacial overdeepening, sometimes favoured by the presence of faults, as in the case of Cregüeña lake. In some cases, they were filled with sediments and developed into peat bogs.

Other erosional landforms are glacial thresholds, characterised by smooth, elongated elevations. Overdeepening basins (Figure 3c), which have circular shapes with scarps over 20 m high, usually contain lakes at their bottoms. Glacial scarps (<10 m) have linear or curved shapes indicating the direction of ice flow.

U-shaped landforms are preserved in the Ésera, Vallibierna, Cregüeña and Salenques valleys. At certain points in the Ésera valley and on the northern slope of the Salenques and Cregüeña areas, two U-shaped levels are preserved, separated by a 200 m difference in altitude, indicating the occurrence of at least two significant glacial episodes. Hanging valleys such as Alba or Forau Tancau valleys (Figure 3a) are tributaries of the Ésera valley.

4.1.3. Periglacial landforms

Periglacial features began to develop during the last deglaciation phase, during a warm period during MIS3 (Vidaller et al., 2024) and especially after the Oldest Dryas cold period (18.5–15.4 ka BP; Crest et al., 2017; Vidaller et al., 2024). Their formation intensified in parallel with glacier shrinkage and today represents the second most common group of landforms. This category is characterised by talus slopes (90% of the group; Figure 2b). Above 2800 m a.s.l., the mean annual ground temperature is below 0°C, indicating the presence of permafrost and active periglacial processes (Serrano et al., 2019). Notable periglacial features are observed in the Alba cirque, including a debris lobe (2760 m a.s.l.), a protalus lobe within the potential permafrost belt, and a rock glacier (García et al., 2017; González-García et al., 2011; Serrano et al., 2019). The thawing of permafrost and the increase in talus slopes and rockfalls increase the risks for mountaineers and climbers in these mountains. The Maladeta massif is home to eight relict rock glaciers (Cregüeña rock glacier is highlighted in Figure 3f). Four of them are located near cirque walls, where they are protected from solar radiation and constantly supplied with rockfalls. The others, located near the Alba lakes, the Salenques valley, the Alba cirque and below the Vallibierna col, flow directly from the walls of the glacial cirques. Of these eight rock glaciers, the one in the Alba cirque is probably the only active one (Serrano et al., 2011).

Protalus ramparts are preserved in the upper elevations of the Maladeta massif, located very close to the walls of the glacial cirques. At present, there is no snow on any of these protalus ramparts all year round.

4.1.4. Fluvial landforms

Fluvial landforms predominate in the valley bottoms, as small rivers that contribute to erosion processes. Ravines located on the slopes merge into the valley floor. Floodplains typically occur in areas where the river has anastomosing or meandering features. The alluvial fans described have a triangular shape and form in areas where the ravines extend into the valley bottoms. Old channels refer to erosive forms that represent dry ravines. These landforms are consistently found on limestone, which is due to the interaction between fluvial and karst processes.

4.1.5. Karstic landforms

The karstic landforms are visible in the limestones in the bottom of the headwaters of the Ésera valley (Figure 1b). Ponors and springs play an important role in altering the course of rivers and ravines, especially during periods of low water flow, such as the dry season (summer). Notably, a significant part of the water from the northern slopes of the NE of the massif bypasses the Ésera river and instead flows into the Garonne river via the Aiguallut ponor (sinkhole).

Sinkholes typically have alignment patterns (Figure 3b) that follow the natural flow of the water. Occasionally, these sinkholes are filled with till deposits. Poljes are sometimes filled with fine sediments reworked from glacier deposits. Most of the karren are interpreted as rillenkarren.

4.1.6. Hillslope landforms

The slope features were created by the debuttressing on the valley walls after the formerly large glaciers had retreated to the cirques. Landslides are observed in low-lying areas and affect Devonian shales and limestones. In contrast, debris flows are more common at higher elevations, often associated with the presence of talus slopes and facilitated by water flows. These features are closely linked to deglaciation processes and are still very active today.

4.2. Thickness of the LIA moraines as an indicator of glacier development

During the LIA, there were a total of ten glaciers in the Maladeta massif, arranged from west to east as follows: Alba, Maladeta, Aneto, Barrancs and Tempestades on the northern slope; Cregüeña, Coronas, Llosas and Russell on the southern slope; and Salenques with an eastern orientation (Figure 4).

Figure 4. LIA moraines thickness. Light purple corresponds to lower values, while dark purple is associated with the greatest thickness. In some cases, the moraine margins have been simplified to obtain a more accurate thickness data.

This map shows LIA moraine thickness of the ten glaciers. Increase in darkness implies higher moraine thickness until a maximum of 91 m.

The moraines thickness associated with Alba (31 m) and Russell (15 m) glaciers are the thinnest, reflecting the limited development of these glaciers during the LIA. Conversely, moraines over 80 m in height are associated with Maladeta (82 m), Llosas (88 m), and Cregüeña (91 m) glaciers (Figure 4).

North-facing glaciers, the largest of the massif, generally have less moraine thickness than their south-facing counterparts, but are larger and extend to lower elevations. Frontal moraines are consistently overridden by proglacial fans. Interestingly, there is no discernible correlation between moraine thickness and glacier area; rather, it depends on the excavation of the cirques. This fact shows that the largest glaciers are not always thickest (Table 1). For example, the Cregüeña glacier, with a relatively modest area of 19.6 ha during the LIA, formed the thickest moraine at 91 m. In contrast, the Aneto glacier, the largest one (249.6 ha), developed a maximum moraine height of only 51 m. This could be due to the fact that the south-facing glaciers excavated more their cirques during earlier glacial periods, and that the thickness of the LIA glaciers depends on the depth of the cirques, so that the south-facing glaciers have a greater thickness than the other north-facing glaciers, even though they are smaller.

Table 1. Mean characteristics of the LIA glaciers and its moraines.

	Orientation	Mean glacier elevation (m)	LIA area (ha)	Maximum moraine thickness (m)	Cirques slope (%)
Alba	NE	2945	12.4	31.2	38.4
Maladeta	NE	2890	120.6	82.4	39.4
Aneto	NE	2926	254.4	51.0	40.0
Barrancs–Tempestades	NE	2770	131.3	38.0	48.3
Salenques	E	2835	57.6	76.0	44.0
Cregüeña	SW	3062	19.6	91.2	43.1
Coronas	SW	2935	41.8	41.8	38.6
Llosas	SW	3054	37.6	88	49.4
Russell	SW	3003	11.5	15	36.5

The mean elevation of the glacier has been calculated with the centroid of the polygon that defined the glacier. The mean slope of the cirques has been calculated only considering the slope between the LIA moraines and the cirque’s walls. This value represents the overdeepened of the cirques.

4.3. Shrinkage of the last remaining glaciers since 1956

Since the LIA, the Pyrenean glaciers have retreated almost continuously from an estimated initial area of 2060 ha (610 ha in the Maladeta massif; Rico et al., 2017). The process of retreat and mass loss has accelerated in recent years, particularly in terms of ice thickness (Izagirre et al., under review; Vidaller et al., 2021, 2023).

In 1956, the glaciated area in the Maladeta massif amounted to 242.4 ha, distributed over eight glaciers (Alba, Maladeta, Aneto, Coronas, Barrancs, Llosas, Tempestades and Salenques), as shown in Table 2. By 1990, the two smallest glaciers (Llosas and Alba glaciers) had disappeared, resulting in a 16% decrease in glacierised area to 203.5 ha (−1.14 ha per year). The trend continued, and by 2000 the glaciated area had further diminished to 181.1 ha (−1.39 ha per year since 1956).

Table 2. Area (ha) of the individual glaciers for the years 1956, 1990, 2000, 2011, 2020 and 2023.

		1956	1990	2000	2011	2020	2023
Alba		0.8	–	–	–	–	–
Maladeta	West	59.8	54.3	48.8	5.3	4.3	–
	East				25.5	20.8	13.2
Aneto	Principal	114.4	100.6	90.9	62.7	44	34
	Secondary					3.8	–
Coronas		12.5	8.3	6.6	1.6	–	–
Barrancs		16	13.7	11.5	6.6	2.6	–
Llosas		2.1	–	–	–	–	–
Tempestades		28.3	19.8	17.4	7.9	6.2	4.7
Salenques		8.5	6.8	5.9	–	–	–
Total area (ha)		242.4	203.5	181.1	109.6	81.7	51.9

The area data for 2011, 2020 and 2023 are taken from Vidaller et al. (2021, 2023) and Izagirre et al. (under review).

In 2011, the remaining glaciers began to separate into smaller bodies (Maladeta glacier divided in two bodies) (Figure 5), resulting in a remaining ice-covered area of 109.6 ha, a significant decrease of 54.8% (−2.41 ha per year) compared to 1956. During the period 2000–2011, the Salenques glacier degraded to an ice-patch. In 2020, there were still six glaciers (equivalent to four glaciers in 1956, Aneto glacier has divided in two bodies in 2016) with an area of 81.7 ha, a decrease of 66.3% compared to 1956. The annual rate of area loss remained constant at −2.51 ha per year (Table 2).

Figure 5. Area of individual glaciers from 1956 to 2023. The polygons defining the shape of the glaciers are colour-coded red for year 1956, blue for year 1990, yellow for year 2000, purple for year 2011, orange for year 2020 and green for year 2023. The outlines for 2011, 2020 and 2023 are from Vidaller et al. (2021, 2023) and Izagirre et al. (under review).

The shrinkage of the glaciers of Maladeta massif since 1956 has been divided in six stages: year 1956 (in red), 1990 (in blue), 2000 (in yellow), 2011 (in purple), 2020 (in orange) and 2023 (in green).

In 2023, the losses intensified and only 51.9 ha of glacier ice remained, a significant decrease of 78.6% compared to the 1956 ice mass (Table 2). The annual rate of area loss accelerated to −2.81 ha per year, leaving only three very small glaciers (Aneto, Maladeta East and Tempestades).

5. Conclusions

We have presented a detailed geomorphological map of the Maladeta massif at a resolution of 1:15,000. The map illustrates the predominant influence of glacial and periglacial processes in this area, but also shows notable examples of fluvial, karstic and hillslope processes.

The investigation of the thickness of the LIA moraines provides valuable insights into the historical state of the LIA glaciers. In particular, our results show no discernible correlation between moraine thickness and the area covered by glaciers. Instead, they suggest that glacier thickness depends on the intensity of cirque carving and the orientation of the glaciers.

In addition, the reconstruction of the glacier surface from 1956 to 2023 provides a tool for understanding the rapid changes that have taken place in the high-mountain landscape, especially in recent years. This comprehensive study contributes to a more complete understanding of the geomorphologic evolution of the Maladeta massif, and the detailed map presented here serves as a basis for future studies.

Software

The geomorphological map of the Maladeta massif was digitally created using ArcGIS 10.5.1 and ArcGIS Pro 3.0.3 softwares, from ESRI. This tool was also used to calculate the percentage of area covered by each geomorphological feature and to determine the LIA moraine thickness for each glacier. Final refinement of the map, including the legend, toponyms, map title, authors, and affiliations, was performed using Adobe Illustrator CC 2014. The LIA moraine map was created using the same programmes, ArcGIS 10.5 and Adobe Illustrator CC 2014.

Acknowledgments

This research was made possible through the support of the Interreg-POCTEFA project OPCC ADAPYR, as well as the projects MARGISNOW (ref PID2021-124220OB-I00) and SNOWDUST (ref TED2021-130114B-I00). Ixeia Vidaller acknowledges financial support from the grant FPU18/04978 and is currently enrolled in the PhD programme at the University of Zaragoza. We extend our gratitude to Chemary Carrera, Guillermo Pérez, Marcel Galofré and Francisco Rojas, for their invaluable assistance during the fieldwork. Additionally, we would like to express our appreciation to AEMET for providing climatic data from the Renclusa and Besurta stations, to Marco from Clima y Nieve Pireneo for providing climatic data from Llanos del Hospital station, and to the Posets-Maladeta Natural Park for providing climatic data from Aneto station and for allowing the field work permits necessary to this investigation.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, Ixeia Vidaller, upon reasonable request.

Additional information

Funding

This research was made possible through the support of the Interreg-POCTEFA project OPCC ADAPYR, as well as the projects MARGISNOW (ref PID2021-124220OB-I00) and SNOWDUST (ref TED2021-130114B-I00). Ixeia Vidaller acknowledges financial support from the grant FPU18/04978.