The higher elevation forests of Norfolk Island are regularly immersed in the clouds and scientific and anecdotal evidence suggests that in addition to rainfall, water is likely to be collected as cloud droplets are intercepted by the forest canopy. This water is likely to be important for the local hydrology and ecology, yet it has never been quantified. To address this, a field measurement campaign was established to measure hydrological inputs to the forest floor at two elevated forest sites in the Norfolk Island National Park. Instrumentation included throughfall and stemflow systems and measurements of rainfall in the open in nearby clearings. Sites exhibited very high stem density and basal area by international standards and delivery of water to the forest floor was dominated by stemflow because of the funnelling characteristics of the dominant palm and pine trees. Both sites showed similar hydrological behaviour with stemflow and throughfall of around 48% and 32%, respectively. Stemflow contributions of 48% far exceed observations from the literature which are typically less than 10%. Rainfall rarely occurred in the absence of low-level cloud and some cloud immersion events lasted for many days with hydrologic inputs continuing for extended periods despite rainfall not being observed in the open. Cloud interception accounted for approximately 20% of total water input at both sites which is equivalent to 25% extra water on top of rainfall measured in the open. From an island-wide perspective the calculated extra hydrological input is only small due to the limited spatial extent of elevated forest, however, the additional water is likely to be very important to local hydrological processes and the unique plants, insects and animals which inhabit the higher elevation forests of Norfolk Island.

Cosmic-ray neutron sensors (CRNS) have been used in many studies for measuring soil moisture and snow pack over intermediate scales. Corrections for geomagnetic latitude, barometric pressure and atmospheric humidity are well established, however, corrections for the effect of solar activity on neutron count rates have been overly simplistic, typically relying on one neutron monitor station and accounting for latitude and elevation crudely or not at all. Recognizing the lack of a generalised and scientifically robust approach to neutron intensity correction, we developed a new approach for correcting CRNS count rates based on analysis of data from 110 quality-controlled neutron monitor stations from around the world spanning more than seven decades. Count rates from each monitor were plotted against the count rates from Climax, CO, USA or Jungfraujoch, Switzerland depending on the time period covered. Relationships between relative counting rates at the site of interest versus the reference neutron monitors were found to be strongly linear. The dimensionless slope of this linear relation, referred to as τ, was shown to increase with increasing geomagnetic latitude and elevation. This dependence of τ on geomagnetic latitude and elevation was represented using an empirical relationship based on a single reference neutron monitor. This generalised approach enables τ to be derived for any location on Earth and also lends itself to roving CRNS studies. The correction procedure also includes a location-dependent normalisation factor which enables easy substitution of an alternative reference neutron monitor into the correction procedure.