A hydrological framework for persistent river pools

8 Persistent surface water pools along non-perennial rivers represent an important water resource for 9 plants, animals, and humans. While ecological studies of these features are not uncommon, these are 10 rarely accompanied by a rigorous examination of the hydrological and hydrogeological characteristics 11 that create or support the pools. Here we present an overarching framework for understanding the 12 hydrology of persistent pools. We identified perched water, alluvial through flow and groundwater 13 discharge as mechanisms that control the persistence of pools along river channels. Groundwater 14 discharge is further categorized into that controlled by a geological contact or barrier (not previously 15 described in the literature), and discharge controlled by topography. Emphasis is put on clearly defining 16 through-flow pools and the different drivers of groundwater discharge, as this is lacking in the literature. 17 A suite of diagnostic tools (including geological mapping, hydraulic data and hydrochemical surveys) 18 is generally required to identify the mechanism(s) supporting persistent pools. Water fluxes to pools 19 supported by through-flow alluvial and bedrock aquifers can vary seasonally and resolving these inputs 20 is generally non-trivial. This framework allows the evaluation of the susceptibility of persistent pools 21 along river channels to changes in climate or groundwater withdrawals. Finally, we present three case 22 studies from the Hamersley Basin of north-western Australia to demonstrate how the available 23 diagnostic tools can be applied within the proposed framework. 24 hydrology is typically poorly understood, and the treatment of the hydrology of persistent river pools

discharge into a stream channel over a low-permeability geological layer caused by the reduced the 225 vertical span of the aquifer (Fig. 3a); where this vertical span reduces to zero is known colloquially as 226 the aquifer "pinching out". This mechanism has been identified as driving regional groundwater 227 discharge to streams (Gardener et al., 2011), but to our knowledge has not yet been explicitly discussed 228 in the context of persistent river pools. 229 Outflow of groundwater where a catchment is constrained by hard-rock ridges that constrict 230 groundwater flow (by reducing the lateral span of surface flow and the aquifer) can also support the 231 persistence of surface water pools (Fig. 3b). Although the importance of catchment constriction has 232 been identified by practitioners (e.g. Queensland Government, 2015), to our knowledge the discharge 233 of groundwater caused by catchment constriction as a mechanism for surface water generation has not 234 previously been described in published literature (springs or otherwise). For example, pools that are supported by the discharge of deep regional groundwater are potentially 285 vulnerable to groundwater abstraction, while perched pools are unlikely to be impacted. Thus, if 286 managing impacts from groundwater abstraction, then monitoring efforts would be best directed to the 287 groundwater-dependant pools at the expense of pools that are disconnected from the groundwater 288 system. It is also important to note the potential logistical constraints that can apply when installing any 289 infrastructure for sampling and monitoring in-stream pools. Persistent pools in arid landscapes are 290 commonly sites of environmental and cultural significance (Finn and Jackson, 2011;Yu, 2000) so that 291 appropriate approvals and permissions typically must be obtained prior to the installation of monitoring 292 infrastructure. This may restrict the types of data that can be collected. Moreover, some sites may be 293 sacred sites, limiting who is able to access them. Surface water features in general are a draw for 294 travellers and roaming livestock, so that any infrastructure must be secure from theft or damage. Flood 295 events and sudden, flashy streamflows are also potential threats to infrastructure, with substantial 296 sediment and vegetation (branches, trees) transported across the floodplain to heights of 2-3 m that can 297 (and have) destroyed sampling equipment. Furthermore, because regional groundwater inputs can be a 298 relatively small (but important) component of the water balance of pools, snapshot sampling commonly 299 targets the end of the dry season. This is when the contribution of regional groundwater is likely to be 300 at its greatest. However, when un-seasonal or early rainfall occurs, or if infrastructure has been 301 damaged, that endpoint in the water balance may not be captured. 302

303
Landscape position can provide some clues as to the mechanism controlling the persistence of a given 304 pool. For example, a pool located high in the catchment on impermeable basement rock is likely to be 305 a perched pool. A pool that is immediately prior to a ridge that constrains the catchment is likely to be 306 supported by geologically constrained groundwater discharge. Lateral catchment constriction can 307 commonly be identified from publicly available aerial imagery, but identification of vertical catchment 308 constriction will usually require geological data from drilling or regional-scale geophysical surveys. it is important to understand that regional-scale groundwater maps are always based on point-data of 335 hydraulic heads measured in the groundwater system, interpreted by a hydrogeologist in the context of 336 what is known about geology and surface drainage (Siegel, 2008). These maps can be refined based on 337 measures of groundwater salinity and groundwater residence times (from environmental tracer data), 338 both of which generally increase along a groundwater flow path. As such, these maps are limited by the 339 spatial distribution of the data available (commonly sparse) and therefore may not accurately capture 340 local-scale features and processes relevant to a particular pool of interest. Nevertheless, if an interpreted 341 water table surface suggests that the regional water table is tens of meters below ground in the vicinity 342 of a pool, then the surface water is likely (but not definitely) perched. If a pool is situated in a region 343 that has been identified as a regional groundwater discharge zone, then this groundwater discharge is 344 likely to be supporting pool persistence. 345 If instrumentation can be installed in the pool, then it may be possible to characterize the pool water 346 balance. Once a pool becomes isolated from the flowing river, and in the absence of rainfall, a general 347 pool water balance is given by; 348 where is the volume of water in the pool (L 3 ), is time (T), is the water flux from the subsurface 350 into the pool (L 3 T -1 ), is the water flux out of the pool into the subsurface (L 3 T -1 ), is the evaporation 351 rate (L T -1 ) and is the surface area of the pool (L 2 ). The water level in the pool, hp (L), can be routinely 352 measured by installing pressure transducers, but conversion of water levels to pool water volume 353 requires knowledge of pool bathymetry, and the relationship between hp and V will change during the 354 dry season as the pool water level recedes, or if pool bathymetry is altered by scour and/or sediment 355 deposition during flood events. Evaporation rates can be taken from regional data or empirical hydraulic mechanism(s) supporting pool persistence (Table 1). For perched pools, which are 362 disconnected from the groundwater system, = = 0, so that the only component of the water balance 363 is water loss through evaporation. Pools that are supported by alluvial through-flow are hydraulically 364 connected to the water stored in the streambed alluvium. Water levels within this alluvium will be more 365 dynamic than regional groundwater levels, so that influx and efflux rates that can change over time in 366 response to rainfall events or seasonal drying (of the near-subsurface). For pools supported by 367 groundwater discharge, influx will dominate over efflux ( > Q o ). If the groundwater discharge is 368 over an impermeable aquiclude (see Fig. 3b) there will commonly be a seepage zone up-gradient of the 369 pool so that water influx is via surface inflow, but outflow to the subsurface can form a source of 370 groundwater recharge to the adjacent (down-gradient) aquifer. If the groundwater discharge is 371 controlled by topography, then the pool will be a site of regional groundwater discharge so that local 372 groundwater recharge (and o ) should be negligible. 373 If a pool is connected to the groundwater system (or o ) can be estimated from Darcy's Law; 374 where is hydraulic conductivity, ∆ℎ ∆ is the hydraulic gradient between the pool and the source aquifer, 377 and i is the area over which the groundwater inflow occurs (which will usually be less than the total 378 area of the base of the pool). The major limitations of this approach are that K of natural sediments 379 varies by ten orders of magnitude (Fetter, 2001), and that the area of groundwater inflow needs to be 380 assumed or estimated using a secondary method. Hydraulic gradients between pools and streambed 381 sediments can be measured using monitoring wells or temporary drive points, with Δh usually on the 382 order of centimetres at most. Determination of the hydraulic gradient between regional aquifers requires 383 that the water level in the pool has been surveyed to a common datum and there is a monitoring well 384 near the pool to measure the groundwater level relative to that datum. In shallow, groundwater 2) Topographically controlled seepage from regional aquifer

509
In this section we demonstrate the application of this framework to persistent river pools in north-west 510 Australia. We begin by providing an overview of our understanding of the hydrology of persistent river-511 pools in the Hamersley Basin region. We then present three case studies to demonstrate how some of 512 the tools described in Section 3 can be applied to identify the key hydraulic mechanisms supporting 513 pool persistence, and the implications for pool susceptibility. 514

515
The Hamersley Basin has an arid-tropical climate with a wet season from October to April and a dry 516 season from May to September (Sturman and Tapper, 1996). Average annual rainfall is less than 300 517 mm yr -1 with most rain falling between December and April (www.bom.gov.au). Annual rainfall 518 statistics can vary dramatically, depending on the influence of thunderstorms and cyclone activity. 519 Thunderstorm activity is commonly highly localised, limiting the potential for spatial interpolation of 520 data from individual monitoring sites. Annual evaporation is around 3000 mm yr -1 (www.bom.gov.au), 521 or about ten times annual rainfall, so that permanent surface water is rare. Ranges, spurs, and hills are 522 separated by broad alluvial valleys with numerous deep gorges created by differential erosion. During hydrogeological mechanisms in the framework outlined in Section 2 (Fig. 5). A sub-set of these (22 534 pools) have been investigated in more detail (Fig. 6) to characterize their mode of occurrence 535 (Dogramaci, 2016). Based on data from this subset of pools, we have generalized the distribution of the 536 hydrogeologic mechanisms supporting pool persistence across this landscape (Fig. 7). Perched pools 537 are generally found in elevated, hard-rock areas where erosion has created a deep pool that is shaded to 538 minimise evaporation. For example, there are approximately 20 pools that reside within the ephemeral 539 drainage lines of the Western Range that flow for a few days in response to rainfall; a subset of pools 540 that are deeply incised and shaded persist all year round (those that are shallower and more exposed to 541 sunlight dry out faster and are not perennial). These pools are important ecologically (supporting bat 542 populations) and culturally (supporting traditional hunting practices). Because these pools are not 543 connected to groundwater, they are not directly at risk of depletion by groundwater withdrawals. 544 However, they are susceptible to changes in streamflow that reduce the water storage in the pools at the 545 commencement of the dry-season, either due to reduced inflows or in-filling by sedimentation. 546 Persistent pools that are connected to groundwater are also abundant across the Basin, with the folded 547 and tilted layered sedimentary sequence resulting in numerous exposures of geological contacts at the 548 land surface. Groundwater discharge from the unconfined aquifer through contact springs is therefore 549 a common mechanism supporting persistent river pools in this region. These are particularly prevalent 550 at the intersection of fluvial deposits and erosion-resistant, low permeability basement rocks. 551 Groundwater-fed pools are also present due to catchment constraints where erosion-resistant layers 552 form ridges in the landscape. Pools supported in part (or completely) by alluvial through-flow are also 553 common along the stream channels due to the storage capacity of the coarse alluvial sediments. The

Case Studies 575
The following three case studies demonstrate the application of this framework to three different pools 576 (or pool systems) within the Hamersley Basin. To the best of our knowledge these pools have not been 577 impacted by human activities. These case-studies demonstrate the application of key methods to infer 578 hydraulic mechanisms supporting pool persistence, and the complexity of applying these methods in 579 real-world situations. We start with a simple case, and build complexity with each case study, using 580 data that highlight the temporal and spatial variability in pool hydrochemistry and provide valuable 581 insight into the supporting hydraulic mechanisms (but also limits the appropriateness of basing an 582 assessment on a small number of samples). The implications of these mechanisms for the susceptibility 583 of the pools to groundwater withdrawals or changing climate are also discussed. 584