Caves provide early warning of unprecedented decrease in rainfall recharge of groundwater

doi:10.21203/rs.3.rs-1556439/v1

PDF

Article

Caves provide early warning of unprecedented decrease in rainfall recharge of groundwater

https://doi.org/10.21203/rs.3.rs-1556439/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted 02 May, 2022

You are reading this latest preprint version

Billions of people worldwide rely on groundwater. As rainfall in many regions in the future is projected to decrease, it is critical to understand the impacts of climate change on groundwater recharge. In this study, five caves record a consistent response to a sustained decrease in rainfall across southwest Australia that began in the late 1960s, characterised by a pronounced increase or ’uptick’ in dripwater and speleothem oxygen isotopic composition (δ18O). It is demonstrated that the uptick is in response to the shallow karst aquifers becoming disconnected from recharge due to regional drying. Our findings imply that rainfall recharge to groundwater across this region is no longer reliably occurring. Examination of the longer speleothem record shows that this is unprecedented over at least the last 800 years. A global network of cave dripwater monitoring would serve as an early warning of reduced groundwater recharge elsewhere, while evidence for upticks in speleothem paleoclimate records would provide a longer-term context to evaluate if current groundwater recharge changes are outside the range of natural variability. This study also validates speleothems as recorders of past hydroclimate via amplification of the δ18O signal by karst hydrology highlighting that speleothem δ18O are records of recharge, rather than a direct proxy for rainfall.

Groundwater use is accelerating globally due to population growth^1,2. Further reductions in water availability in many regions due to climate change^3,4 are likely placing increasing pressure on groundwater resources and groundwater dependent ecosystems^5,6. It is critical to understand how drying in the mid-latitudes will impact rainfall recharge to groundwater resources globally. Currently our understanding of the sensitivity of groundwater systems to climate change is limited^4,7,8. This is because water table monitoring using direct methods such as bores can be complicated by groundwater extraction and land use change which can obscure the climate change signal⁹, and the water table may be recharged by rivers or lakes¹⁰ in addition to rainfall recharge. Furthermore, the hydraulic memory of fossil groundwater systems can delay a response to climate change by decades or longer¹. These factors mean that attributing changes in groundwater recharge to the impacts of climate change on rainfall is difficult.

Caves offer a natural ‘window’ to observe the movement of water from the surface through the vadose zone and into groundwater aquifers, and can thus indicate rainfall recharge to groundwater^11,12. Rainfall recharge may be focussed along fractures, fracture zones and conduits, or it can be diffuse porous flow where highly permeable host rock are present. In karst environments recharge is likely to be a mixture of the two flow types^11,13. Similar to groundwater, the oxygen isotopic composition (δ¹⁸O) of cave dripwaters and consequently speleothems, is largely determined by the recharge-weighted δ¹⁸O, particularly in water-limited climates¹⁴ where groundwater resources are increasingly under pressure. Accordingly, speleothems offer a highly temporally-resolved record of rainfall recharge to groundwater. Modern speleothems formed since the early 20^th century can be used to assess the response of rainfall recharge to groundwater during recently observed climate change, while older speleothem material can provide a longer-term context of the range of natural, pre-industrial variability of recharge to groundwater. As speleothems form in the vadose zone above the water table, they are not complicated by modern groundwater abstraction nor the longer term processes such as changes in hydraulic boundary conditions.

Rainfall in southwest Australia is highly seasonal and falls in the cooler months, typically from May to October, owing to the seasonal migration of the Southern Hemisphere westerly winds. The region has experienced a prolonged decline in annual rainfall attributed to changes in Southern Hemisphere atmospheric circulation systems with a significant contribution from anthropogenic forcing^15,16. In response to anthropogenic changes in ozone and then greenhouse gases¹⁶ rainfall has decreased by 10-25% overall across the region since the 1960s¹⁷. Several local studies have since recognised the effect of reduced rainfall and increased abstraction on water resources within southwest Australia, as direct rainfall recharge has generally been found to be the major recharge mechanism^18,19. For example, the Rottnest Island freshwater lens (Fig. 1) has decreased in size by 40% primarily due to the rainfall decrease¹⁸. Likewise, water table decline within the Perth Basin has been identified as a result of decreasing rainfall^19,20. However, quantifying the amount of water table decline attributable to decreased rainfall input is complicated by increasing groundwater extraction over the same time period²¹. Climate models indicate, with high confidence, that rainfall in this region is likely to decline further throughout the 21st century¹⁶ which will continue to impact water availability in southwest Australia^19,20. This will have a consequent impact on water-dependent ecosystems²² and water security for this region that currently depends on groundwater to meet 75% of its water demand¹⁹.

In this study we constructed a regional speleothem δ¹⁸O record using eight speleothems from five caves across a 300 km transect in southwest Australia (Fig. 1). The stalagmites have all grown through the period of regional rainfall decline. The speleothem δ¹⁸O records, along with cave dripwater monitoring, demonstrate a common rise, or ‘uptick’, in values beginning in the 1990s. This change is not driven by a change in the input source water (rainfall) δ¹⁸O values but rather is demonstrated to be a shift in the hydraulic flow properties in the karst aquifer in response to reduced rainfall recharge. We argue that the declining rainfall has reduced rainfall recharge to groundwater along focussed flow paths. We also assess this in the context of the longer (paleo) speleothem record for the region and demonstrate that this reduction in recharge is unprecedented for the last 800 years. This is the first study to use speleothems from multiple caves to examine the response of reduced rainfall recharge due to climate change in the modern record. In doing so, it also confirms that ‘upticks’ in speleothem δ¹⁸O values primarily represent reduced water supply to the speleothem. This substantiates the interpretation of speleothem δ¹⁸O datasets as a proxy for hydroclimate, and importantly validates the hydrological flow path response to rainfall changes control on recharging water δ¹⁸O values.

Regional speleothem δ¹⁸O record

Eight speleothem records from five caves, as well as cave dripwater δ¹⁸O data from one of the caves (Fig. 2) were used to construct a regional speleothem δ¹⁸O record (Fig. 3a). These records all demonstrate a consistently replicated rise in δ¹⁸O values during the late 20^th Century and early 21^st Century. The beginning of this rise varies between speleothem records, with the rise beginning as early as the late 1960s in one speleothem record, i.e., MND-S1 (Fig. 2). The variability in the timing of the δ¹⁸O response could be due to different hydraulic response times (i.e. cave flow properties) or speleothem age uncertainty. Nevertheless, the composite data indicate that the isotopic rise becomes a common signal by the 1990s. The speleothem records show that this rise is approximately 1‰, while our dripwater data confirm that this rising trend has continued to the present indicating an overall increase of 1.5‰ above the 1900-1968 mean speleothem δ¹⁸O value (Fig. 3a).

Speleothem δ¹⁸O is a function of multiple processes, including: i. cave temperature; ii. possible reaction processes that can result in speleothem δ¹⁸O being precipitated ‘out of isotopic equilibrium’ with its source water; iii. rainfall δ¹⁸O; and iv. impacts of flow-path variability on dripwater δ¹⁸O^23,24. Yet despite the possibility for multiple processes, the distinct and replicated large rise in speleothem δ¹⁸O values indicates a common response to forcing that is occurring across the region (Fig. 3a). The regional uptick in speleothem δ¹⁸O cannot be attributed to the temperature-dependent fractionation that occurs during calcite precipitation²³ as recent warming¹⁶ would lower speleothem δ¹⁸O by an estimated −0.27 ‰ below pre-1960 values rather than raise speleothem δ¹⁸O values as observed (Fig. 3a). Enhanced ‘out of isotopic equilibrium processes’^23,25 were shown not to be dominating d¹⁸O variability in the modern Golgotha Cave records (representing four of the eight stalagmites in our composite)²⁶. Furthermore, neither temperature nor calcite precipitation disequilibrium could produce the rising trend observed in the dripwater δ¹⁸O values. The rising dripwater δ¹⁸O trend that is concurrent with and extends the speleothem δ¹⁸O trend, confirms that the speleothem δ¹⁸O values are recording a primary signal arising from the dripwater δ¹⁸O. This confirms that rainfall δ¹⁸O and/or flow path variability must be the common driver of the regional uptick in δ¹⁸O values.

Rainfall recharge and the speleothem δ¹⁸O response

Rainfall δ¹⁸O (the source of dripwater δ¹⁸O) is considered to be the primary driver of dripwater δ¹⁸O variability¹⁴. Long-term monitoring of rainfall δ¹⁸O in southwest Australia²⁷ can constrain the impact of climate change on rainfall δ¹⁸O in the region. The decline in southwest Australian winter rainfall has occurred in two steps¹⁷ occurring in 1968/1969 and 1999/2000 (Fig. 3b). This pattern of rainfall decline is also evident over the narrow coastal limestone belt where the caves are located (Fig. 3b). The rainfall decrease has been characterised as a reduction in total rainfall and heavy rainfall during winter, as well as a decrease in the total number of rain days²⁸. Of these three factors, it is rainfall intensity that most impacts the δ¹⁸O composition of southwest Australian rainfall²⁷. However, using both measured and modelled rainfall δ¹⁸O values, we can quantify that this accounts for between a −0.4 ‰ to +0.1 ‰ change in rainfall δ¹⁸O over the period of rainfall decline (Table 1), whereas the rise in cave δ¹⁸O data is fifteen times greater (Fig. 3a). Shallow groundwater extracted from the freshwater lens of Rottnest Island, for which mean δ¹⁸O values are −4.0 ‰ (Fig. 3c) and generally reflect mean δ¹⁸O composition of the recharging rainfall¹⁸, further supports the relatively small change in rainfall δ¹⁸O. Thus, we argue that the overall rise of 1.5 ‰ in our regional δ¹⁸O record reflects a rise in dripwater δ¹⁸O despite little change in the δ¹⁸O of annual rainfall supplying these drips.

A recent study identified that rainfall recharge through Golgotha Cave occurs via both diffuse flow through the matrix and focussed flow along fractures²⁶. Importantly, it was also shown that dripwater from fracture dominated flow paths is up to 1 ‰ lower than the dripwater supplied by diffuse flow²⁶. The isotopic distinction is due to activation of flow along fractures preferentially recharging karst stores with water that is relatively ¹⁸O-depleted as it is sourced from higher intensity rainfall²⁶. The isotopic distinction between flow types is seen in the distribution of δ¹⁸O values for diffuse and fracture flow dripwaters (Fig. 4). The distributions of diffuse flow drips overlap with that for all daily rainfall events ≥ 2 mm, whereas fracture flow drip δ¹⁸O values are lower and overlap with the distribution of daily rainfall when recalculated to only include daily totals for heavy (≥ 5 mm) rainfall events. This is characteristic of the typical recharge-bias response seen at a global scale in cave drips^14,26. Figure 4 also shows that the distribution of fracture flow δ¹⁸O values closely agrees with that of Holocene groundwater values supporting that rainfall recharge to groundwater beneath these caves occurs primarily via fracture flow.

The eight speleothems from five different caves in our composite record will each have unique flow path properties and length, yet all exhibit rising d¹⁸O values, suggesting a common hydrological response to the regional decline in rainfall. A reduction in the contribution of focused flow along fractures supplying drips is the only mechanism that can explain the magnitude of the δ¹⁸O uptick in the regional speleothem δ¹⁸O record. A reduction in focussed flow will also led to a volumetric decline in rainfall recharge to groundwater and this is observed in the local water table beneath the cave systems (Fig. 3d). Three caves in the region have water table level measurements that span wholly or partially over the period of the rainfall decrease. The longest record that began in the 1950s confirms a decline in the volume of rainfall recharge to the local water table that began around 1980 (Fig. 3d), in agreement with the beginning of the speleothem δ¹⁸O uptick. A further two records of cave water table measurements that commenced in the late 1990s also support further large drops in the water table below at least two other caves in the region (Fig. 3d).

Movement of water focused along fractures will occur and be enhanced once overlying soils have become saturated²⁹, consistent with observations at Golgotha Cave³⁰. A reduction in soil moisture and/or the length of the season of soil saturation will reduce the overall contribution of recharge from focused flow, and consequently impact the δ¹⁸O values of the percolating water. We examined temporal variability in soil moisture by extracting modelled values for the five cave sites for 1911-2019³¹ (see Methods and Supplementary Fig. 2). A decline in modelled soil moisture is evident over the same period as the δ¹⁸O uptick (Fig. 3a,e). Prior to 1969, mean values across the region for the number of days/year when soils were saturated was 96 (Fig. 3e). This decreased to 67 days/year following 1969 and an overall decrease of 42% to 55 days/year after 1999 CE when the δ¹⁸O uptick is also steepest (Fig. 3a,e). A long-term reduction in soil moisture further supports our interpretation that the mechanism of reduced focused flow is driving the uptick in the cave δ¹⁸O record.

The rise in regional speleothem δ¹⁸O values accelerates in the last two decades (Fig. 3a). Thus, it appears that the majority of the speleothem δ¹⁸O increase is in response to the second rather than the first rainfall decrease. This likely reflects the cumulative impact of reduced rainfall on soil moisture, whereby reduced saturation in the soil layers reduces the likelihood of focussed flow along fractures occurring, and the duration that it occurs for each wet season. That is, we propose that fracture activation is still occurring but was reduced during the period of the first rainfall decrease (1969 to 1999), whilst fracture activation was further reduced to infrequent or no fracture activation from the commencement of the second rainfall decrease after 1999. A further consideration is the lag for karst stores to drain. For example, fracture activation evident at one site that was monitored in Golgotha Cave took at least three years for dripwater δ¹⁸O to return to values equivalent to matrix-flow drips²⁶. Thus the impact of a reduced water balance has decreased recharge along focussed flowpaths in the karst and led to the ‘disconnection’ of karst aquifers from rainfall recharge, resulting in the rise in dripwater and speleothem δ¹⁸O values.

Implications for understanding rainfall-recharge to groundwater

The replication of the uptick signal across the southwest Australian region validates the interpretation that the contribution of diffuse versus focussed flow to dripwater dominates stalagmite δ¹⁸O²⁶, as no other mechanism could produce this common response across multiple caves. Our study offers a robustly constrained processed-based relationship between rising speleothem δ¹⁸O values in response to a well-documented and sustained decrease in rainfall. Importantly, it has been shown that this is due to the impact of varying rainfall recharge to the soil and the karst rather than a shift in rainfall d¹⁸O, highlighting that this response amplifies the speleothem d¹⁸O signal. In the global literature, there are examples of similar isotopic upticks in speleothem paleoclimate records, typically prior to a hiatus in speleothem growth^32,33. Thus, identification of replicated upticks in speleothems can be used to identify past reductions in rainfall recharge to groundwater in a region.

The speleothem record from Golgotha Cave²⁶ provides a longer-term context to assess the recent d¹⁸O uptick against. Stalagmite GL-S4 grew continuously since 1200 CE with a robust chronology constructed from annual laminae counting (Fig. 5a). It is fed by diffuse dripwater supplied from the matrix but also by focused flow along fractures that activate during wetter years²⁶. Speleothem d¹⁸O values for GL-S4 are −3.9 ‰_VPDB or lower, consistent with fracture flow for the entire record until the mid-1990s, when higher values indicate that diffuse flow dominates (Fig. 5a,b). Figure 5a,c compares the speleothem GL-S4 recharge d¹⁸O record with the Law Dome ice core accumulation-based rainfall reconstruction (see Fig. 1 for location) for southwest Western Australia^15,34. The two records are in agreement in terms of the unprecedented nature for the last eight centuries of both the recent rainfall decrease depicted in the Law Dome record and the corresponding reduced rainfall recharge in the speleothem record. We confirm that no replicated uptick events are seen in the last 800 years in the stalagmite record from the southwest Australian region²⁶ suggesting that the current decoupling of rainfall to groundwater recharge is unprecedented for the last eight centuries. This has important implications for contextualising the threat to the biodiversity in groundwater dependent ecosystems and sustainable use of groundwater in a region heavily dependent on it for agriculture, and where future use is anticipated to increase²⁰.

In this study, caves located in the vadose zone and where recharge can be identified unambiguously from rainfall, have provided an early warning of decreased rainfall recharge to groundwater resources. The recent speleothem and dripwater dataset indicates that rainfall recharge to groundwater may no longer be reliably occurring, while the longer speleothem record demonstrates that this is unprecedented for at least the last 800 years and highlights the immediate threat of climate change to water security in a region heavily dependent on groundwater. This is of relevance for future climate change as there is debate over the modelled increase in rainfall intensity and whether that would lead to increased hydrological responses. Runoff (i.e. flooding) has been the focus of research and adaptation planning so far³⁵, with little attention until now given to the implications of future climate change on groundwater recharge. This study shows that a network of monitored caves in other regions where groundwater is relied upon could serve as an early warning of reduced recharge, particularly in mid-latitude regions where widespread rainfall declines are expected this century as a result of human-caused climate change³⁶. Monitoring also provides valuable data for ground-truthing speleothem paleoclimate records, while upticks in the speleothem paleoclimate records provide a longer-term context for evaluating the risks to groundwater resources from natural and human-caused climate changes.

Speleothem chronologies. Chronologies for all speleothems were published previously. Briefly, all stalagmites had been actively forming at the time that they were removed from the caves and the date of removal from the cave is used as the youngest age. The chronologies for CRY-S1³⁷, YD-S2³⁸ GL-S3 and GL-S4²⁶ were determined from annual laminae counting of 2D maps of Sr concentration and/or visible layering formed from fabric/colloidal matter in thin section; MND-S1 from annual laminae counting of Ba concentration from duplicate transects³⁹; LAB-S1 from the peak fitting method of ref. ⁴⁰ utilising principal components analysis of duplicate transects of Ba, Sr and U concentrations; while GL-S1 and GL-S2 chronologies were determined from ‘bomb pulse’ detection using radiocarbon analyses⁴¹ and U/Th disequilibrium dating²⁶ for details of chronologies for Golgotha Cave stalagmites.

Speleothem δ¹⁸O. Data were previously published for all records^26,38,42 with the exception of CRY-S1 and LAB-S1. Milling and isotope ratio mass spectrometry methods are as described in ref. ²⁶. CRY-S1 d¹⁸O measurements were made at the Mark Wainwright Analytical Centre at UNSW Sydney using a MAT-253 isotope ratio mass spectrometer with Kiel carbonate device (25–50 μg samples). LAB-S1 d¹⁸O measurements were made at the Research School of Earth Sciences, ANU, using a Thermo MAT-253 isotope ratio mass spectrometer coupled to a Kiel IV carbonate device (110–130 μg samples). Data in both laboratories are normalised to the Vienna Pee Dee Belemnite (VPDB) scale using NBS19 (δ¹⁸O = −2.20 ‰ and δ¹³C = +1.95 ‰) and NBS-18 (δ¹⁸O = −23.0 ‰ and δ¹³C = −5.0 ‰).

Golgotha Cave dripwater and farmed calcite δ¹⁸O. Dripwater data, sampled at 4-6 weekly intervals; and farmed calcites, in situ for 9-31 months, were previously published in ref. ²⁶. Data and method for converting dripwaters from the VSMOW scale to a calcite equivalent on the VPDB scale were also previously published²⁶ using mean annual cave temperature measured at Golgotha Cave (14.6 ±0.1°C) and the bulk empirical fractionation factor for Golgotha Cave (1.0318 ±0.0004) determined therein.

Composite speleothem record. Speleothem, and dripwater δ¹⁸O data (as calcite equivalent values) were interpolated to be annually resolved to construct the composite record. For Golgotha Cave speleothem records this represented a 3-5x increase in resolution, for YD-S2 a 1.2x increase in resolution; and for records CRY-S1 and LAB-S1, this represented a 2-3x decrease in resolution relative to the original data. Alternative methods were investigated such as maintaining the sampled resolution and weighting each data point according to the time interval that it represents, and the uptick in δ¹⁸O values is still robust i.e., the choice of method for compositing the data did not affect the interpretation. We considered uncertainty in the age-depth models but this is least at the youngest end. i.e., 0% at the year of collection and still small at 1900 CE (range ±3 to 12%). It was not necessary to account for analytical uncertainty in δ¹⁸O as these are small (0.05 ‰ for carbonates and 0.1 ‰ for water)²⁶.

Rainfall δ¹⁸O data. Annual rainfall amount for southwest Australia (line joining 30^oS, 115^oE and 35^oS, 120^oE) and the extent of the Tamala Limestone (Fig. 1) was extracted from Bureau of Meteorology gridded spatial data³¹.

Annual and monthly amount weighted δ¹⁸O was measured at the Perth Global Network of Isotopes in Precipitation (GNIP) site⁴³ and Calgradup²⁷.

Predicted amount weighted annual δ¹⁸O calculated from an empirical relationship between daily δ¹⁸O and daily precipitation, Eq. (3) in ref. ²⁷, applied to Bureau of Meteorology rainfall observations from the ‘high quality’⁴⁴ sites Boyanup and Cape Naturaliste.

Soil moisture data. Modelled soil moisture data were extracted for each cave site from the Australian Bureau of Meteorology’s Australian Water Resources Assessment Landscape model: AWRA-L version 6⁴⁵. Briefly, AWRA-L is a continent-wide water balance model with 0.05° gridded outputs of soil moisture, runoff, evapotranspiration and drainage. We utilised daily soil moisture in the lower (0.1-1 m; term ‘Ss’) and deep soil store (1-6 m; term ‘Sd’), as well as rainfall (term P_g) and actual evapotranspiration (term E_tot)⁴⁵. Model output (Ss) is validated against field measurements of soil moisture at Golgotha Cave⁴⁶ (see Supplementary Figure 2) and the Sd output, corresponding to the observed depth of seasonal wetting prior to initiation of percolation into the host rock³⁰, is used to calculate the number of days per year when soil is saturated. The best match between observed and modelled soil moisture data is 66%; however, observed recharge at Golgotha suggests a higher percentile (78^th) would be representative of fracture activation necessary for recharge and we apply this to all 5 sites. See Supplementary Figure 2 for further details.

ONLINE SUPPORTING MATERIAL

Supporting figures, tables, text and associated references, are contained in the Supplementary Information.

DATA AND CODE AVAILABILITY STATEMENT

The Golgotha speleothem and drip water time series data may be found on the PANGEA paleoclimate data repository (PDI-30276 and PDI-30277). The authors agree to archive the LAB-S1, CRY-S1 and YD-S2 records on the PANGEA paleoclimate data repository upon the manuscript being accepted for publication (https://www.pangaea.de/).

Weather station and gridded climate data are available from the Australian Bureau of Meteorology, http://www.bom.gov.au.

ACKNOWLEDGEMENT

This research is supported by funding from the Australian Research Council (DP140102059 to PCT and NJA). The Sr map used for CRY-S1 chronology was obtained on the X‐ray fluorescence microscopy (XFM) beamline at the Australian Synchrotron, part of ANSTO under proposals PA14312. The authors would like to thank Carolina Paice (Parks and Wildlife Services at Leeuwin Naturaliste National Park) and Mark Delane (Margaret River Busselton Tourism Association) for providing cave water level data and reports. We would also like to thank Nevena Kosarac and Liza McDonough for assistance with laminae counting of CRY-S1, and Lewis Adler and Mark Wainwright for IRMS analytical support. This research was undertaken with the assistance of resources and services from the National Computational Infrastructure (NCI), which is supported by the Australian Government. IsoGSM data which appears in this article was provided thanks to Kei Yoshimura.

AUTHOR CONTRIBUTING STATEMENT

This manuscript was devised by PCT, AB and SCP. SCP, PCT and ADG collated data and prepared the figures. PCT, NJA, and KTM conducted analyses and provided data. SCP wrote the manuscript with input from all co-authors.

1. Cuthbert MO, Taylor RG, Favreau G, Todd MC, Shamsudduha M, Villholth KG, et al. Observed controls on resilience of groundwater to climate variability in sub-Saharan Africa. Nature 572(7768), 230-234 (2019).

2. Jasechko S, Perrone D, Befus KM, Bayani Cardenas M, Ferguson G, Gleeson T, et al. Global aquifers dominated by fossil groundwaters but wells vulnerable to modern contamination. Nat Geosci 10, 425 (2017).

3. Hughes CE, Cendón DI, Johansen MP, Meredith KT. Climate Change and Groundwater. In: Jones JAA (ed). Sustaining Groundwater Resources. (Springer, Dordrecht, 2011).

4. Cuthbert MO, Gleeson T, Moosdorf N, Befus KM, Schneider A, Hartmann J, et al. Global patterns and dynamics of climate–groundwater interactions. Nat Clim Chang 9(2), 137-141 (2019).

5. Dragoni W, Sukhija BS, London GSo. Climate Change and Groundwater. (Geological Society, 2008).

6. Kløve B, Ala-Aho P, Bertrand G, Gurdak JJ, Kupfersberger H, Kværner J, et al. Climate change impacts on groundwater and dependent ecosystems. J Hydrol 518(PB), 250-266 (2014).

7. Hughes JD, Petrone KC, Silberstein RP. Drought, groundwater storage and stream flow decline in southwestern Australia. Geophys Res Lett 39(3), (2012).

8. Taylor RG, Scanlon B, Döll P, Rodell M, van Beek R, Wada Y, et al. Ground water and climate change. Nat Clim Chang 3, 322 (2012).

9. Cochand F, Brunner P, Hunkeler D, Rössler O, Holzkämper A. Cross-sphere modelling to evaluate impacts of climate and land management changes on groundwater resources. Sci Total Environ 798, (2021).

10. Wöhling T, Wilson S, Wadsworth V, Davidson P. Detecting the cause of change using uncertain data: Natural and anthropogenic factors contributing to declining groundwater levels and flows of the Wairau Plain aquifer, New Zealand. Journal of Hydrology: Regional Studies 31, (2020).

11. Mahmud K, Mariethoz G, Baker A, Treble PC, Markowska M, McGuire E. Estimation of deep infiltration in unsaturated limestone environments using cave lidar and drip count data. Hydrol Earth Syst Sci 20, 359–373 (2016).

12. Beal LK, Wong CI, Bautista KK, Jenson JW, Banner JL, Lander MA, et al. Isotopic and geochemical assessment of the sensitivity of groundwater resources of Guam, Mariana Islands, to intra- and inter-annual variations in hydroclimate. J Hydrol 568, 174-183 (2019).

13. Jukić D, Denić-Jukić V. Groundwater balance estimation in karst by using a conceptual rainfall-runoff model. J Hydrol 373(3-4), 302-315 (2009).

14. Baker A, Hartmann A, Duan W, Hankin S, Comas-Bru L, Cuthbert MO, et al. Global analysis reveals climatic controls on the oxygen isotope composition of cave drip water. Nature Communications 10(1), 2984 (2019).

15. Zheng Y, Jong LM, Phipps SJ, Roberts JL, Moy AD, Curran MAJ, et al. Extending and understanding the South West Western Australian rainfall record using a snowfall reconstruction from Law Dome, East Antarctica. Clim Past Discuss 2020, 1-32 (2020).

16. Delworth TL, Zeng F. Regional rainfall decline in Australia attributed to anthropogenic greenhouse gases and ozone levels. Nat Geosci 7(8), 583-587 (2014).

17. Hope P, Timbal B, Fawcett R. Associations between rainfall variability in the southwest and southeast of Australia and their evolution through time. International Journal of Climatology 30(9), 1360-1371 (2010).

18. Bryan E, Meredith KT, Baker A, Post VEA, Andersen MS. Island groundwater resources, impacts of abstraction and a drying climate: Rottnest Island, Western Australia. J Hydrol 542, 704-718 (2016).

19. Ali R, McFarlane D, Varma S, Dawes W, Emelyanova I, Hodgson G, et al. Potential climate change impacts on groundwater resources of south-western Australia. J Hydrol 475, 456-472 (2012).

20. McFarlane D, Stone R, Martens S, Thomas J, Silberstein R, Ali R, et al. Climate change impacts on water yields and demands in south-western Australia. J Hydrol 475, 488-498 (2012).

21. Barron O, Froend R, Hodgson G, Ali R, Dawes W, Davies P, et al. Projected risks to groundwater-dependent terrestrial vegetation caused by changing climate and groundwater abstraction in the Central Perth Basin, Western Australia. Hydrological Processes 28(22), 5513-5529 (2014).

22. Barron O, Silberstein R, Ali R, Donohue R, McFarlane DJ, Davies P, et al. Climate change effects on water-dependent ecosystems in south-western Australia. J Hydrol 434-435, 95-109 (2012).

23. Lachniet MS. Climatic and environmental controls on speleothem oxygen-isotope values. Quat Sci Rev 28(5), 412-432 (2009).

24. Fairchild IJ, Smith CL, Baker A, Fuller L, Spötl C, Mattey D, et al. Modification and preservation of environmental signals in speleothems. Earth-Sci Rev 75(1), 105-153 (2006).

25. Deininger M, Hansen M, Fohlmeister J, Schröder-Ritzrau A, Burstyn Y, Scholz D. Are oxygen isotope fractionation factors between calcite and water derived from speleothems systematically biased due to prior calcite precipitation (PCP)? Geochim Cosmochim Acta 305, 212-227 (2021).

26. Treble PC, Baker A, Abram NJ, Hellstrom JC, Crawford J, Gagan MK, et al. Ubiquitous karst hydrological control on speleothem oxygen isotope variability in a global study. Communications Earth & Environment 3(1), 29 (2022).

27. Griffiths AD, Treble P, Hope P, Rudeva I. Rainfall Stable Water Isotope Variability in Coastal Southwestern Western Australia and Its Relationship to Climate on Multiple Timescales. J Geophys Res Atmos 127(1), 31 (2022).

28. Philip P, Yu B. Interannual variations in rainfall of different intensities in South West of Western Australia. International Journal of Climatology 40(6), 3052-3071 (2020).

29. Ireson AM, Butler AB, Wheater HS. Evidence for the onset and persistence with depth of preferential flow in unsaturated fractured porous media. Hydrology Research 43(5), 707-719 (2012).

30. Leopold M, Gupanis-Broadway C, Baker A, Hankin S, Treble P. Time lapse electric resistivity tomography to portray infiltration and hydrologic flow paths from surface to cave. J Hydrol 593, 125810 (2021).

31. Bureau of Meterology. Climate Data Online. 2020 [cited 2020 August 2020]http://www.bom.gov.au/climate/change/index.shtml#tabs=Tracker]. Available from: http://www.bom.gov.au/climate/change/index.shtml#tabs=Tracker

32. Asrat A, Baker A, Leng MJ, Hellstrom J, Mariethoz G, Boomer I, et al. Paleoclimate change in Ethiopia around the last interglacial derived from annually-resolved stalagmite evidence. Quat Sci Rev 202, 197-210 (2018).

33. Treble PC, Baker A, Ayliffe LK, Cohen TJ, Hellstrom JC, Gagan MK, et al. Hydroclimate of the Last Glacial Maximum and deglaciation in southern Australia's arid margin interpreted from speleothem records (23-15 ka). Clim Past 13(6), 667-687 (2017).

34. van Ommen TD, Morgan V. Snowfall increase in coastal East Antarctica linked with southwest Western Australian drought. Nat Geosci 3, 267 (2010).

35. Fowler HJ, Wasko C, Prein AF. Intensification of short-duration rainfall extremes and implications for flood risk: current state of the art and future directions. Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences 379(2195), 20190541 (2021).

36. IPCC. Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to

the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. (Cambridge University Press, 2021).

37. Baker A, Mariethoz G, Comas-Bru L, Hartmann A, Frisia S, Borsato A, et al. The Properties of Annually Laminated Stalagmites-A Global Synthesis. Reviews of Geophysics 59(2), e2020RG000722 (2021).

38. McDonough L, Treble P, Baker A, Borsato A, Frisia S, Nagra G, et al. Past fires and post-fire impacts reconstructed from a southwest Australian stalagmite. Earth ArXiv, (2021).

39. Treble P, Shelley JMG, Chappell J. Comparison of high resolution sub-annual records of trace elements in a modern (1911–1992) speleothem with instrumental climate data from southwest Australia. Earth Planet Sci Lett 216(1), 141-153 (2003).

40. Nagra G, Treble PC, Andersen MS, Bajo P, Hellstrom J, Baker A. Dating stalagmites in mediterranean climates using annual trace element cycles. Scientific Reports 7(1), 621 (2017).

41. Markowska M, Fohlmeister J, Treble PC, Baker A, Andersen MS, Hua Q. Modelling the ¹⁴C bomb-pulse in young speleothems using a soil carbon continuum model. Geochim Cosmochim Acta 261, 342-367 (2019).

42. Treble P, Chappell J, K. Gagan M, McKeegan K, Mark Harrison T. In situ measurement of seasonal δ¹⁸O variations and analysis of isotopic trends in a modern speleothem from southwest Australia. Earth Planet Sci Lett 233, 17-32 (2005).

43. Hollins SE, Hughes CE, Crawford J, Cendón DI, Meredith KT. Rainfall isotope variations over the Australian continent – Implications for hydrology and isoscape applications. Sci Total Environ 645, 630-645 (2018).

44. Lavery B, Joung G, Nicholls N. An extended high-quality historical rainfall dataset for Australia. Australian Meteorological Magazine 46(1), 27-38 (1997).

45. Frost AJ, Ramchurn A, Smith A. The Australian Landscape Water Balance model (AWRA-L v6). Technical Description of the Australian Water Resources Assessment Landscape model version 6: Bureau of Meteorology Technical Report; 2018).

46. Treble PC, Bradley C, Wood A, Baker A, Jex CN, Fairchild IJ, et al. An isotopic and modelling study of flow paths and storage in Quaternary calcarenite, SW Australia: implications for speleothem paleoclimate records. Quat Sci Rev 64, 90-103 (2013).

47. Yoshimura K, Kanamitsu M, Noone D, Oki T. Historical isotope simulation using Reanalysis atmospheric data. J Geophys Res Atmos 113(D19), (2008).

48. Priestley SC, Meredith KT, Treble PC, Cendón DI, Griffiths AD, Hollins SE, et al. A 35 ka record of groundwater recharge in south-west Australia using stable water isotopes. Sci Total Environ 717, 135105 (2020).

Table 1: A comparison of annual rainfall δ¹⁸O values before and after southwest Australian rainfall decreased. 1968/1969 was used for the timing of the decrease following the analysis of ref. ¹⁷.

Dataset	Measured and modelled mean annual rainfall δ¹⁸O values (‰)
Dataset	1908-1968	1969-2018	Difference
Perth measured⁴³	−3.7 ±0.3 (1962-1966)	−4.1 ±0.5 (1973-1976, 1984-1995, 1998-1999, 2006-2014)	−0.4
Calgardup measured²⁷	-	−4.4 ±0.5 (2006-2018)	-
Boyanup simulated²⁷	−4.5 ±0.2	−4.4 ±0.2	+0.1
Cape Naturaliste simulated²⁷	−4.1 ±0.2	−4.1 ±0.2	0
IsoGSM⁴⁷	-	−4.2 ±0.3 (1979-2018)	-

There is NO Competing Interest.

d18OuptickSupplementary.docx
Caves provide early warning of unprecedented decrease in rainfall recharge of groundwater

PDF

Version 1

posted 02 May, 2022

You are reading this latest preprint version

Caves provide early warning of unprecedented decrease in rainfall recharge of groundwater

Status:

Version 1

Abstract

Figures

Introduction

Results And Discussion

Regional speleothem δ¹⁸O record

Rainfall recharge and the speleothem δ¹⁸O response

Implications for understanding rainfall-recharge to groundwater

Methods

Declarations

References

Tables

Competing Interests

Supplementary Files

Status:

Version 1

Caves provide early warning of unprecedented decrease in rainfall recharge of groundwater

Status:

Version 1

Abstract

Figures

Introduction

Results And Discussion

Regional speleothem δ18O record

Rainfall recharge and the speleothem δ18O response

Implications for understanding rainfall-recharge to groundwater

Methods

Declarations

References

Tables

Competing Interests

Supplementary Files

Status:

Version 1

Regional speleothem δ¹⁸O record

Rainfall recharge and the speleothem δ¹⁸O response