*2.3. Shoreline Change Data*

The primary data used to document changes in position of the Napatree barrier were historical high-water line shoreline positions [24,25,63] (Table 2). Shorelines were derived from T-Sheets (1883 and 1948), georeferenced vertical aerial photographs (1939, 1951, 1963, 1975, 1985 and 1992) and digital orthophotographs (2004, 2012 and 2014). Shoreline positions and annualized end-point rates using previously published shorelines [24,25,63] were reanalyzed at transects spaced 50 m alongshore relative to an onshore baseline. Reanalysis involved the inclusion of the 1883 and 1948 T-Sheets, which were not used in the original study by Boothroyd et al. [23], and examining individual shoreline pairs (e.g., 1883 and 1939; 1939 and 1948). Shoreline positional uncertainty was assessed in the original mapping by Hapke et al. (2010); Boothroyd and Hehre (2007); and Boothroyd et al. (2016). Total positional uncertainty (Up) of a shoreline at each transect was calculated in these studies as the square root of the sum of five relevant uncertainty terms squared (Equation (1) for NOS T-Sheets; Equation (2) for historical aerial photographs)

$$\mathbf{U\_{P}} = \sqrt{(\mathbf{U\_{g}}^{2} + \mathbf{U\_{d}}^{2} + \mathbf{U\_{t}}^{2} + \mathbf{U\_{pd}}^{2})} \tag{1}$$

$$\mathbf{U\_P} = \sqrt{(\mathbf{U\_g}^2 + \mathbf{U\_d}^2 + \mathbf{U\_a}^2 + \mathbf{U\_{pd}}^2)}\tag{2}$$

**Figure 5.** (**A**). Oblique aerial photograph of the Napatree Barrier and Sandy Point Barrier taken in 1936. Note the beach cottages and the presence of dune vegetation along the barrier. (**B**). Oblique aerial photograph of the Napatree Barrier taken on 24 September 1938, 3 days after the 1938 hurricane made landfall. North is to the left of both images, (**A**,**B**). Note the inlet at the east (far) end of the barrier and numerous surge channels across the barrier. (**C**). An oblique aerial photograph taken after Hurricane Carol (1954) of the eastern end of the NPCA (note the two groins installed between 1948 and 1951) (**D**). Car buried in a washover fan on the east end of the Napatree Barrier during Hurricane Carol (1954). This photograph is just out of frame to the east of Figure 5C. See Figure 1 for the location of the photograph [62]. View is towards the east.

**Table 2.** NOS T-Sheets (NOS-T), aerial photographs (AP) and digital orthophotos (DO) used in this study. Dates with an asterisk denote that a dataset was used to digitize a shoreline position. Time since storm indicates the number of days between the photograph/survey and last time the threshold water level (0.7 m MHHW) was exceeded at the Newport tide gauge.


These relevant terms include: georeferencing uncertainty (Ug), digitizing uncertainty (Ud), T-sheet uncertainty (Ut), aerial photograph uncertainty (Ua) and uncertainty in the position of the high-water line (Upd) [25]. The total positional uncertainties for these shorelines are reported in Table 3. The uncertainty of the end point change between two shorelines (Upair) was taken the square of the sum of positional uncertainty of the two shorelines (Equation (3)), and these were calculated as annualized rates (Ur) by dividing that results by the number of years between shoreline surveys (Equation (4)). Linear regression was applied to the shoreline positions for comparison with the long-term (1883–2004) shoreline change rates of Hapke et al. [23]. Barrier width was measured at each transect as the straight-line distance between the last high-tide swash shoreline of the bayside and ocean sides of the barrier.

$$\mathbf{U\_{pair}} = \sqrt{\left(\mathbf{U\_{p1}}^2 + \mathbf{U\_{p2}}^2\right)} \tag{3}$$

$$\mathbf{U\_r = U\_{pair}/(year\_2 - year\_1)}\tag{4}$$

**Table 3.** Reported measurement and total positional uncertainties reported for the shorelines used in this analysis. Sources notes by the superscript next to the date. <sup>1</sup> Hapke et al. [25]. <sup>2</sup> Boothroyd and Hehre [24]. <sup>3</sup> Boothroyd et al. [23].

