| |
| ---------- | ||
| Home | Research | Publications |
| ---------- |
Although seasonal variations in temperature are
relatively small in Hawaii, rainfall is highly seasonal. Highest
rainfall occurs in winter months when the North Pacific subtropical
high pressure ridge shifts southward, allowing penetration of North
Pacific storm systems as far south as the islands. The abundance of
rainfall in Hawaii also varies on interannual and interdecadal time
scales, mainly because of the influences of ENSO events and the NPI
[Chen, 2003; Chu, 1989]. Precipitation in Hawaii is represented by the
Hawaiian rainfall index (HRI). HRI was constructed using the
precipitation data from nine stations on each of the three islands
(Kauai, Oahu, and Big Island). The stations represent
rainfall from varying elevations and varying locations with respect to
the trade wind direction.
Significant decadal variability in the atmosphere over the North Pacific has been found in the intensity of the wintertime Aleutian low. The strength of the Aleutian low is represented by the North Pacific Index, which is defined as the area-weighted mean sea level pressure (mb) over the region 30N-65N, 160E-140W [Trenbeth and Hurrel, 1994] (http://www.cgd.ucar.edu/~jhurrell/np.html). NPI has a period of around 20-30 years. The 20th century experienced 4 main NPI cycles. Negative NPI periods are identified by a strong Aleutian low. The SST in the central Pacific is also lowered during that period. Positive NPI regimes prevailed between 1900-1922 and again between 1948-1976, while negative NPI regimes dominated between 1923-1947 and from 1977 to the late 90's [Deser et al., 2004].
Variations in the Aleutian low have a clear signature on Hawaiian
precipitation. For the period 1906-2001 the NPI was the primary
forcing of Hawaiian rainfall. A negative (positive) NPI is correlated
with low (high) precipitation in Hawaii [Chen, 2003; Deser et al.,
2004]. Moreover, the Aleutian Low is a major forcing of the North
Pacific and has great societal impacts on North America.
ENSO events currently recur on a time scale of 2-7 years, but ENSO frequencies, duration and intensities have likely varied over time [Tudhope et al., 2001; Wang and An, 2001]. During warm ENSO events, the North Pacific atmospheric high shifts closer to the islands leading to abnormally dry weather in Hawaii [Chu, 1989; Lyons, 1982]. Winter rainfall records from Hawaii correlate well with records of ENSO events over the past 100 years [Schroeder, 1993]. Many factors affect rainfall abundance in Hawaii, and not all periods of drought are correlated with warm ENSO events. It is clear, however, that warm ENSO events are nearly always associated with low levels of winter rainfall. The effects of ENSO events in Hawaii are nonlinear: warm ENSO events tend to induce dry conditions whereas there is a poor correlation between Hawaiian precipitation and cold ENSO events [Chu, 1989].
ENSO events are identified here using the Niņo 3.4 index [Barnston and
Chelliah, 1997]. Trenbeth's definition of the Niņo 3.4 index
(http://www.cgd.ucar.edu/cas/catalog/climind/TNI_N34/index.html) is
that an El Niņo (La Niņa) occurs if 5-month running means of sea
surface temperature (SST) anomalies in the Niņo 3.4 region (5N-5S,
120-170W) exceed 0.4 degC (-0.4 degC) for 6 consecutive months or
more. ENSO warm (cold) events correspond to El Niņo (La Niņa).
Variations in mass, energy, and momentum resulting from a redistribution of equatorial rainfall regimes due to ENSO events are communicated into mid- to high latitude [Webster, 1994]. So, even if NPI is the primary driver of Hawaiian rainfalls variations, this effect has the potential to extend ENSO influence beyond the Tropics and modulate variations in the Aleutian low. Masuda [2002] suggested that NPI represents 2 stable states in which the transitions could be triggered by ENSO events. Teleconnections between ENSO and the Pacific Decadal Oscillations
(PDO) have also been suggested [Chen, 2003; Gershunov et al., 1999;
McCabe and Dettinger, 1999]. Newman et al. [2003], however provided
evidence for the PDO being driven by ENSO and Schneider and Cornuelle
[2004] have shown that the PDO is in fact a response to the North
Pacific Index (NPI), ENSO and the Kurishio current extension. As the
PDO is not a "primary driver", in this study we will mainly discuss
the effects of the NPI and ENSO on Hawaiian rainfall.
As lake Waiau is at an elevation of 3970m, it is exposed to the dry atmosphere far above the NE trade wind inversion. Although the rainfall amount is less above the inversion layer than beneath it, the temporal variations are the same and are influenced by ENSO and NPI. Indeed, Mauna Loa Slope Observatory (MLSO) precipitation variations are correlated with the HRI with 95% significance (r=0.73). Moreover, MLSO precipitation displays correlation coefficient of 0.49 and -0.45 respectively with NPI and Niņo 3.4 index for the period 1956-2003 with a 95% significance.
MLSO and Lake Waiau should get approximately the same amount of rainfall as they are at the same elevation (570m difference). Rainfall variations at the summit of Mauna Kea and at MLSO have a correlation coefficient of 0.53, significant at 95% for the period 1973-1981. Thus, the rainfall variations in Lake Waiau should reflect the Hawaiian Islands rainfall variations.