Large-Scale Beach Change: Kaanapali, Hawai'i

Abstract

Using monthly beach profile surveys and historical aerial photographs, the seasonal and long-term (48 year) beach morphology for Kaanapali Beach, Maui is described. By identifying the shoreline position in historical aerial photographs it is determined that the Kaanapali area is subject to long periods of mild erosion and accretion punctuated by severe erosional events related to short-period Kona storms and hurricane waves. Increased Central Pacific tropical cyclone activity of the late 1950's and early 1960's and Hurricane Iniki in 1992 are identified as contributing factors to the observed volume change during these periods. Between these erosional periods the Kaanapali shoreline is relatively stable characterized by light erosion to moderate accretion suggesting the recovery time may be on the order of roughly 20 years. Over the 48-year period 1949 to 1997, the Kaanapali and Honokowai cells have experienced a net sediment volume loss of 43,000 ±730 m3 and 30,700 ±630 m3 respectively for a total net volume loss of 73,700 ± 990 m3. The Kona storms and hurricanes of the early 1960's and 1992 collectively account for 136,000 m3 of sediment lost or approximately 62% of the gross volume change for the entire period, revealing the significant erosional effect of these storms. Recovery after each of these storms accounts for 73,900 m3 or approximately 33% of the gross volume change. A residual loss of 10,600 m3 representing 5% of the gross volume change is inferred as chronic erosion and may be a product of relative sea-level rise (RSLR). An increase in short-period southwesterly wave energy during these erosional periods is well documented and may have transported beach sediment further offshore than normal (beyond the reef) and is identified as a possible mechanism for long-term erosion in this area. The spatial distribution of historical shoreline movement suggests the majority of sediment transport occurs in the central portion of Kaanapali near Kekaa and Hanakoo Point and is driven by longshore rather than cross-shore transport. Surveyed beach profiles reveal a strong seasonal variability with net erosion in the summer and accretion in the winter with an along the shore-alternating pattern of erosion and accretion. 65% of the net volume change occurs south of Kekaa Point confirming the more dynamic nature of the southern (Kaanapali Cell). Net beach profile volume change from the mean suggests that June and January are the most dynamic months each with approximately 14% of the total volume change. We attribute the significant and rapid erosion and accretion events due to wave-induced longshore transport of sediment. Field observations of monthly beach sediment impoundment in the Kaanapali cell are examined and compared to three models that predict longshore sediment transport (LST). Beach profile results indicate sediment impoundment occurs seasonally with a nearly balanced longshore sediment transport system between profile 5 and 9. Longshore transport rates are derived from seasonal cumulative net volume change in the middle of Kaanapali Beach at profile 7. Cumulative net sediment transport rates are 29,379 m3/yr ±15% to the north and 22,358 m3/yr ± 6% to the south for summer and winter respectively, a net annual rate of 7,021 m3/yr ± 10% to the north and a gross annual rate of 51,736 m3/yr ± 2%. Predictive transport formulas such as CERC (1984), CERC (1991) and Kamphius (1991) predict net annual transport rates at 3 x 103 percent, 77 percent and 6 x 103 percent of the observed transport rates respectively. The presence of fringing reef significantly effects the ability of the LST models to accurately predict sediment transport. When applying the CERC (1984, 1991) and Kamphius (1991) formulas, the functional beach profile area available for sediment transport is assumed much larger than actually exists in Kaanapali because of the presence of a fringing reef that truncates a portion of the sandy profile area. The CERC (1984, 1991) and Kamphius (1991) formulas don't account for the presence of a reef system which may contribute to the models overestimate of longshore sediment transport as they assume the entire profile is mobile sediment. However the fact that the CERC (1991) model underestimates the observed transport implies that additional environmental parameters (such as wave height, direction and period) playa more substantial role than the influence of the reef in the model results. The CERC (1991) Genesis model is found to be superior in fitting the observed longshore transport at Kaanapali Beach. The success of the Genesis model is partly attributed to its' ability to account for short-term changes in near-shore parameters such as wave shoaling, refraction, bathymetry, antecedent conditions and several other shore face parameters not accounted for in the CERC (1984) or Kamphius (1991) formulas. The use of the CERC (1984) formula is prone to practical errors in its' application particularly in the use of the recommended "K" coefficient and wave averaging that may a significantly overestimate the LST. A better fit to the observed LST is achieved with the CERC (1984) if the K value is decreased by an order of magnitude from 0.77 to 0.07. The Kamphius (1991) formula is especially sensitive to extremes in wave period and tends to deviate from observed transport estimates for unusually high wave periods (this study) and approximates observations nicely in areas with low wave periods (Ping Wang et al. (1998). Many of the studied predictive LST formulas are prone to overestimate transport and thus their use requires a comprehensive understanding of the complexities and errors associated with employing them. Great care must be used when applying LST models in areas with significant hard bottom or shallow reefs that alter the beach profile shape. Due to these errors, the use of the CERC (1984) and Kamphius (1991) formulas are better suited as a qualitative interpretative tool of transport direction rather than magnitude.