Feature

Weighing Rain

Measuring rainfall involves more than just catching it in a bucket.

Dave Campbell and Earl Bardsley

As the water problems in Auckland have shown, rainfall plays an important part in people's lives. Rainfall measurements were being made in China and India thousands of years ago and from that time on, virtually all rainfall-measuring instruments have been some variation or other on the theme of an open bucket.

Measuring rainfall has always been a cornerstone of the hydrologic sciences, since rainfall forms the input to the land-based portion of the hydrological cycle. Many hydrological problems are tackled using the water balance as a framework, with measurements of precipitation providing a starting point.

Despite advances in weather radar and remote-sensing techniques, most reliable precision measurements of rainfall are made at a single point by traditional rain gauges. These gauges use a tapering funnel of a standard dimension to collect rainwater in a cylinder or bottle, where it can be measured.

There are disadvantages with this approach. Conventional rain gauges cannot account for variations in rainfall over even small distances, especially for short time scales. They can seriously underestimate rainfall in windy conditions, and snowfall cannot be measured reliably.

The hunt for the perfect rain gauge has led us in an unexpected direction recently. We became interested in the response of groundwater systems known as confined aquifers to external forces. It has long been known that changes in barometric, or atmospheric, pressure are reflected by changes in the water levels of open-ended tubes, called piezometers, penetrating these aquifers. This is because part of the force exerted on the ground by the atmosphere is supported by the aquifer's structure while the full load bears down on the piezometer water surface. An increase in barometric pressure thus forces the water level down the tube.

We thought that other types of loading should also affect well water levels, in particular the often-large mass fluxes of water at the land's surface associated with rainfall and evaporation. During a rain storm, for every millimetre of rain which falls, the increase in near-surface mass as a result is equivalent to one kilogram per square metre. This added weight at the surface effectively compresses the aquifer, forcing water into the piezometer and leading to an increase in water level.

There are mechanical factors which make some aquifers potentially more sensitive to this effect. It turns out that aquifers formed in the alluvial sediments of the Waikato and Hauraki are especially sensitive. Here, enormous volumes of sand and gravel were laid down by the historical wanderings of the ancestral Waikato river, interleaved with silts and peat lenses which result in a complex of confined aquifer structures.

At an experimental site at the Matamata aerodrome, we have been measuring hourly water level fluctuations in piezometers penetrating aquifers at depths of 20, 40 and 160 metres, along with rainfall and barometric pressure for about 18 months. Perhaps surprisingly, all three aquifers display an immediate and predictable response to both barometric pressure and rainfall.

The aquifer at 160 metres depth (located in gravels beneath 70 metres of overlying volcanic rock!) responds with the same immediacy as the shallower two. This provides supplementary evidence that all the aquifers are remotely sensing the rainfalls as loading increases, rather than responding directly to the rainfall itself.

There is no possibility that rainfall penetrates rapidly into the ground to recharge the aquifers, and hourly measurements of the water table show no concurrent additions of water from rain. In addition, soil moisture measurements indicate that large amounts of rainfall accumulate near the surface before infiltrating to the water table weeks or even months later.

Huge Rain Gauge

The results of our work have some implications for hydrologists. We have effectively demonstrated for the first time a "rain gauge" capable of measuring precipitation to some precision over an area of hectares to square kilometres. This may allow improved estimation of precipitation in situations where present measurements are inadequate -- for example seasonal snow accumulation in higher latitudes or in forests, which is notoriously variable and difficult to measure with conventional methods.

Ground water scientists have not previously recognised this phenomenon and there appear to be spin-offs in the field. For example, the concept of water recharging of confined aquifers may have to be reconsidered in the light of these results. The textbook confined aquifer receives inputs of water only from the surface, and hydrologists often estimate the recharge rate so that water removal for human use can be allocated in a sustainable way.

The seasonal trends evident for our aquifers would normally be interpreted as recharge, but they are in fact just a loading response to the changes in near surface moisture and so don't represent any "real" water entering the aquifer. Removing water from these aquifers may not be sustainable at all, which has important implications for water resource management.

Dr Earl Bardsley is senior lecturer in hydrology at Waikato University.
Dr Dave Campbell and Dr Earl Bardsley are in the Department of Earth Sciences at Waikato University.