March - 2019

Preliminary Report of the Mulu Caves 2017 Expedition

Summary and results

Mulu Caves 2017 was the 24th Anglo-Sarawak Expedition to Gunung Mulu National Park and operated in the field between 29th March and 27th April 2017. The expedition had several components:

1) A team in the Hidden Valley to continue exploration and surveying of new extensions of Wonder Cave, first discovered via the Conviction Cave entrance in 2015.

2) A team to continue exploration and surveying of the high level Creedence series (discovered in 2014) in Clearwater Cave.

3) A team to look for connections in the Southern Peninsula area of Gunung Api, between Cave of the Winds (part of the Clearwater System), Racer Cave and Lagang’s Cave.

4) A team to climb into a potential high level passage in the roof of Sarawak Chamber.

5) Photography of the expedition and Mulu caves by National Geographic.

6) Scientific study of palaeomagnetic sediments and notch elevation surveys.

In the Hidden Valley, a total 12.5km of new passage was discovered surveyed and photographed over a vertical range of 380m.

In the Clearwater Cave System, 6.8km were added to the surveyed length of the cave system, taking it to a new overall surveyed length of 222km.

1.3km were added to Lagang’s Cave, making it just under 8km total length and 430m were added to Racer Cave, making the total length 6.7km.

A programme of sampling and observational data collection was carried out as part of the ongoing study of the geomorphology and speleogenesis of the Mulu caves.

The team of twenty-eight were led by Andy Eavis and comprised members from the UK, Germany, USA and Belgium plus our hosts at Sarawak Forestry.  The expedition was greatly assisted by National Park Management, staff and local people.

Exploration, Surveying and Photography

Hidden Valley Dave Nixon

Since the discovery of Conviction Cave and its subsequent connection to Wonder Cave during the 2015 reconnaissance expedition to the Hidden Valley, a return visit was inevitable. The way into the eastern side of Gunung Api was openly available and totally unexplored.

A team of eight was carefully selected including members of the previous expedition as well as 4 new but suitably skilled members including a Doctor.

The mobilization to the Hidden Valley was executed by helicopter, hired in Miri and flown via Mulu airport, to collect personnel and approximately 500kgs of equipment. These logistics took meticulous planning and luck had its part to play with the weather for the pre-selected flying day.

On arrival at the Hidden Valley a base camp was again constructed at Prediction Cave and radio communications were established by means of a mast erected on the helipad. In addition, a new satellite phone system was successfully employed as an emergency means of contact. The cave was soon rigged up with approximately 500m of rope and anchor points. This meant the team could base themselves at an underground camp close to the limit of the 2015 exploration and then continue their exploration.

Splitting into multiple small teams, the exploration of the Wonder/Conviction system commenced with progress mainly in a northerly direction towards the Melinau Gorge. In total 12.5km of new passage were discovered, surveyed and photographed over a vertical range of 380m. The limit of known cave passages on the east side of Gunung Api now extends 3.3km north of the Hidden Valley and falls short of exiting in the Melinau Gorge by approximately 2.2km.

The geology of this side of the mountain is strikingly different to that of the Clearwater Cave system and it is now accepted that the passages associated with Wonder/Conviction Cave have been formed by water flowing from the Hidden Valley in a northerly direction. Previously it was commonly assumed that the water flowed south, having sunk in or flowed through the Melinau Gorge area. Surprisingly for the highest cave in Mulu, many active streams were encountered, some of which are quite large and responsible for localised flooding, which was experienced at first hand. Although no dye tests were conducted on this trip, it is thought that this water also flows north to the Melinau Gorge, perhaps to resurge in the upper/eastern side. If so, it may then flow downstream in the gorge to subsequently sink once again into the limestones and flow via the Whiterock River to resurge finally from Clearwater Cave. Future dye traces will be able to shed light on this possibility.

There are also subtle differences in the geology and cave morphology; the Wonder Cave extensions differ from the rest of Mulu, especially in the form of the large scale vadose downcutting. This has left behind some spectacular examples with clear scallops in the deep trench walls that indicate the northerly flow.

Demobilisation from the Hidden Valley was done on foot and all expedition equipment was extracted by porters over a period of 5 days, using a transit camp close to Batu Nigel.

Clearwater Cave – Mark Brown

Creedence Series

The Creedence Series was entered in 2014 via climbs from the Scum ring area of Clearwater Cave; numerous leads were left unexplored, running in all directions. An entrance at the far northerly extent was reached in a large doline (named High Noon) but surface access was not established that year.

The 2017 team located the doline from the survey coordinates and a fly past in the helicopter en route to the Hidden Valley; they subsequently established a route overland to the High Noon entrance. The route goes steeply uphill for 550m, close to Solo Cave, and then traverses rough ground around dolines, enabling the entrance to be reached in approximately 5 hours, slightly quicker than the alternative route through the cave to the northern limit of exploration. From High Noon, four underground camps were used to explore leads in the top level of the Creedence series, as well as one camp at Scum ring to explore leads in the mid-level, “Bladerunner” area, of the series.

One series of passages continued further east beyond the current extent of any known passages at a lower level in that part of Clearwater, reaching a point 1.5km from the Cloud/Cobra/Bridge Cave System, south of the Hidden Valley. Surveys established the highest point in the Clearwater system at 570m above sea level. Several team members traversed the cave from High Noon Entrance to Clearwater Resurgence, a trip that involves a vertical range of 450m, a cave passage age range of around 1.5million years and a spectacular variety of cave landscapes along the way.

A total of 5.3km of new passage was explored, surveyed and photographed in the Creedence Series, adding considerable information on the layout and geomorphology of the area. Palaeomagnetic samples were also collected from several locations in this, the oldest known part of the Clearwater Cave System.

Cave of the Winds – Eureka Chamber

As part of the connection team project (see below) a new chamber was entered from Clay Hall in Cave of the Winds (which is part of the Clearwater System), named Eureka. This chamber was 200m long with spectacular formations, and is geomorphologically significant because it showed that passages could cross the line of the Cave of the Winds River, heading towards Racer Cave. 1.4km were surveyed in passages associated with this Chamber.

Secret Garden

Passages that had been surveyed in the Secret garden area of Clearwater in 2015 were tied in to the Clearwater survey model during the expedition.

Overall, 6.8km were added to the length of the Clearwater Cave system, taking it to a new overall surveyed length of 222km. This maintains its position as 8th longest cave in the world, only 470m behind the Lechuguilla System in the USA. Clearwater is the now the third longest cave system in the world to have been formed by river solution (rather than hypogenic or tidal mixing caves) and probably the largest cave in the world by passage volume.

3) Connection project

The objective of connecting Cave of the Winds (part of the Clearwater System), Racer Cave and Lagang’s cave, all in the southern “peninsula” or “toe” area of Gunung Api, was established after new discoveries in 2014 suggested that the caves were very close and could possibly be linked. The connection would add over 12km in length to the Clearwater System, making it a worthwhile objective. Careful survey work and exploration of numerous passages and aid climbs, many of which were tight, muddy and challenging, were carried out.

Eureka chamber was a significant discovery (see Clearwater description earlier), which brought Cave of the Winds and Racer Cave to within 10m of each other. A sump was discovered at the very end of the expedition, where teams on both sides made radio contact and could hear each other hammering on rock, but no lights could be seen through the sump and its passage was very constricted.

Work in Lagang’s and Racer Cave (via the Easter Cave section surveyed in 2014) also brought those two caves very close together but connections remain elusive. It has been proved that passages in all the caves are the same geomorphologically, but interrupted by calcite or sediment blockages or breakdown. The survey will help direct future attempts at a connection.

In the process of the exploration, 1.3km were added to Lagang’s Cave, making it just under 8km total length. 430m were added to Racer Cave, making the total length 6.7km and 1.4km were added to Cave of the Winds, as described in the Clearwater Cave section above.

4) Sarawak chamber climb

3D laser scanning work in 2015 had suggested the existence of high level passages in the roof of Sarawak Chamber. There was also a conundrum that the size of the current passages entering and exiting this massive chamber seemed too small to explain its formation. A team camped in the chamber and carried out a bolt climb to investigate. The initial route entered an area of poor quality rock, requiring a restart on the opposite wall, which was more overhanging and technical. At a height of 80m above the start point a small passage was found, but it dropped back into the main passage, and was not the large feature that had been hoped for.

5) National Geographic

The National Geographic photographer and journalist were assisted by the expedition members and local assistants to gather a range of material for their assignment.

Preliminary Science reportDr Andrew Farrant.

The 2017 Expedition had several scientific objectives. The first was to refine the chronology of the caves by collecting some additional palaeomagnetic samples from the upper levels of Clearwater that had been discovered in 2014. The second was to get a better understanding of the elevation and gradient of the notches in the southern part of Clearwater by surveying them using laser range finding equipment. The elevation of the notches will be compared with palaeoclimate data derived from speleothems in the Mulu area to see if it is possible to correlate notch formation with specific climatic events. The third objective was to combine speleogenetic observations in the caves (scallops, notches, phreatic and vadose passages) with sediment data to better understand the speleogenesis of the Mulu caves over time.

Palaeomagnetic sampling

Palaeomagnetic samples were collected from 9 sites in Clearwater Cave and Conviction Cave, each site consisting of 6 sediment cores 25 mm long and 25 mm diameter. Four were taken from the Creedence Series, the passages at the highest level in Clearwater Cave, and an additional two from Leap in Time near the Secret Garden. Three samples were collected from Conviction Cave. These will be analysed at the Palaeomagnetic laboratory at the University of Liverpool. However, it is clear that the classic fine grained ‘cricket muds’, associated with the notches in the lower part of Clearwater which were extensively sampled in 1991, are absent or poorly developed in the higher-level passages. In the upper levels, the finer grained sediments were interbedded with gravels, and are not likely to be so good for palaeomagnetic analysis. This may hinder the acquisition of a reliable palaeomagnetic signal from the samples.

Palaeomagnetic samples, Creedence, Clearwater Cave.

Notch Elevations.

The initial assumption was that the notches in the Mulu caves are formed by the throughput of copious amounts of gravel during wetter climatic periods over the last 2.5 million years. This causes the stream to erode laterally rather than incise into the floor of the passage, forming a notch. The advent of better surveying techniques, including the use of laser range-finding devices linked directly to cave surveying software means that accurate elevations and gradients for the notches can be obtained with far greater precision and tied back to fixed surface GPS points. Assuming a constant rate of uplift and valley incision, the age of each notch can be estimated from the palaeomagnetic timescale, and then correlated with global climatic events and climatic data obtained from speleothems in Mulu by Kim Cobb and others.

A preliminary correlation with the Mulu speleothem climate record and notch elevation (=age) suggests there is no clear relationship between notch aggradation and climate. However, fieldwork during this expedition indicates that the notches are far more complex than initially assumed, with some notches having multiple sub-notches, and significant gradients over the length of the Clearwater system. In addition, notches are associated with fan aggradation from both the Melinau Gorge and the Melinau Paku rivers, and so are influenced by different catchment dynamics and responses to climate forcing. There are also likely to be significant lead times from the onset of wetter climatic periods and the aggradation of the cave rivers and notch formation, which is dependent on the distance from the sediment source. This is possibly on the order of several thousand years for the downstream parts of the Clearwater system. Furthermore, aggradation and notch formation in the cave passages is dependent on water from either the Melinau Gorge or the Melinau Paku River being routed through the cave system and bringing in gravels. This is not the case with the present drainage from the Melinau Gorge, where the water comes in via boulder chokes which filter out much of the coarse bed-load. Lastly, whilst the modern DistoX surveys are more accurate than the 1978 cave surveys (from which notch elevations were estimated from cave elevation drawings), there is still some data processing to do with the survey data to get the final elevation for some of the notches. Further work and data processing is required to tease out any correlation between palaeoclimate and notch generation.


During the course of the April 2017 expedition, observations of cave geomorphology and geology were made in parts of Lagang’s Cave, Cave of the Winds, Racer Cave and Clearwater. In addition, the analysis of digital cave survey data viewed in Survex is yielding new insights into the influence of the geological structure on cave formation. Much of the cave passages in Mulu are influenced by geological structure, most notably the strike of certain favourable bedding planes, and two sets of joints, one trending approximately N-S, the other E-W. In the northern part of Gunung Api, the geological structure appears to be relatively simple, with the limestone dipping steeply to the NW at around 60-70°. Most passages in Whiterock are orientated along the strike of the limestone, typically trending at 210-220°. To the south, the strike of the limestone gradually shifts to a more southerly orientation. This is clearly visible in Armistice and Creedence where the passages along strike trend at 195-200°. In the lower levels of Clearwater, the strike is more to the SSE, before swinging back around to a NE-SW trend in the southern end of Gunung Api; here Lagang’s cave trends at 220°. In this part of the cave, the dip is noticeably gentler, at around 25-30°.

Joint patterns are also clearly evident from the cave surveys. The E-W trending joint set is clearly visible in the Armistice and Creedence areas, and in the Secret Garden-Hyperspace-Battleship area. These joints are often used to transfer water laterally from the older beds lower down in the dipping limestone sequence up through the limestone to the younger beds in which the Clearwater resurgence is located. An alternative way for the water to transfer stratigraphically up through the limestone sequence is to descend down-dip and then ascend up a phreatic joint-aligned riser. This is what happened at Deep Thought and Ronnie’s Delight pitches near the Scum ring camp, where water transfers up joints into higher level limestone beds. The same thing happens at Coral Aven in Lagang’s Cave, albeit it at a much smaller scale. In the southern part of Clearwater and the Lagang’s-Racer Cave area, steeply inclined joint sets are often the dominant influence on passage orientation, for example in Fast Lane, Lagang’s Cave.

The exploration of the Hidden Valley cave systems also hints at a structural influence on cave development. The new extensions in Conviction Cave suggest cave development is orientated along the axis of a shallow plunging syncline trending at 220°. Further discussions with the Hidden Valley team and a detailed analysis of the cave survey data are needed.

Sediments and speleothems

During this expedition, several new volcanic ash localities were found, both in Lagang’s Cave and in Creedence, in Clearwater Cave. A potential objective of any future expedition would be to find any speleothem material above or below the ash deposits to better constrain the age of this super-eruption. In addition, some qualitative assessment of the gravels was made in Creedence and Hidden Valley which suggests that the inputs from Hidden valley are dominated by siltstone, unlike the Melinau Gorge which is sandstone dominated. In the Hidden valley, unusual yellow encrustations on cave popcorn and moon-milk deposits were found in Conviction cave, which were collected for analysis. These may be bacterially mediated mineralisation on evaporitic calcite speleothems.

Allophane deposits (weathered volcanic ash) in the Creedence Series.

For the sample list please see the Appendix of this document.


Work in all the components of the expedition has greatly added to the knowledge of the caves, geology and geomorphology of the Gunung Api area. A total of 21km of new cave passage has been surveyed, most of it using modern digital techniques that provide a high degree of accuracy. A large amount of resurvey work has also been carried out to refine the 3D model of the cave systems under Gunung Api.

The different components also had a positive feedback effect; for example, the exploration teams helped collect data for the science and the science work informed the exploration, typified by the Eureka Chamber discovery.

As ever, the expedition came away with a good number of leads and new questions to be answered by future expeditions.

The expedition was also conducted safely, with good emergency planning, and no incidents or accidents were reportable.

A detailed description of all aspects and findings from the expedition will be published as part of the Anglo/Malaysian, Mulu Caves Project 2017 report.


From Sarawak:

The Chief Minister of Sarawak
The Sarawak State Secretary
The Director of Forests, Sarawak
Protected Areas & Biodiversity Conservation Unit, Sarawak Forestry Corporation
Gunung Mulu National Park Management & Staff
Local organisers and assistants
Oswald Braken Tisen
Rambli Ahmad
Veno Enar
Richard Hi (Tropical Adventure)


The expedition is very grateful for the support of the following sponsors and organisations.

National Geographic
Five Ten Ltd.
Spanset Ltd
Ghar Parau Foundation
Mark Wright Training Ltd
Wolfride Clothing

* A full list of acknowledgements will be published in the final report