4.2 Results of the Validation
77
on the modelled oil drift data (Table 4.7). This is mainly due to considerably longer
exposure time of oil in the satellite data than in the modelled oil drift data. Compared
to the performance boundaries, all simulations resulted in mortality in the “within”
category.
4.2.2.2 Marine Mammals
The estimated mortality with ERA Acute for whales in the GOM was considerably
underestimated compared to the field assessment and the performance boundaries.
It is believed that this was primarily due to the physiological factor p phy (probability
of dying given contact with oil film on the sea surface thicker than 10 µm) firstly
being set too low for toothed and baleen whales.
Cetaceans have in general been regarded as little vulnerable towards oil spills and
the original p let factor (p phy × p beh ) was 0.1%, based on early development work for
the ERA Acute model and similar environmental risk analyses methods (e.g. FrenchMcCay 2004, 2009; Østbye et al. 2003; Spikkerud et al. 2004; Spikkerud et al. 2010).
Using the factors from previous work, the highest estimated ERA Acute mortality
for bottlenose dolphins in the Gulf of Mexico using modelled data was 228 individuals and the highest estimate using field data was 1959 individuals, considerably
underestimating the reported mortality.
During more recent development of ERA Acute, the factors were therefore
re-evaluated for toothed and baleen whales (Stephansen et al. 2018) based on
further scientific studies and preliminary reporting of high whale mortality from
the Macondo accident (DHOS) (e.g. Deepwater Horizon Natural Resource Damage
Assessment Trustees 2016). The increase in acute mortality estimated for the
common bottlenose dolphin using the refined factors is illustrated in Fig. 4 12. The
overall mean mortality increases from 1,128 (95% CI = 695–1,708) to 9,796 (95%
CI = 6,256–14,092) individuals.
Currently recommended ERA Acute parameters are presented in Table 4.4 and
are used in the results (Table 4.7), based on these calibrated vulnerability numbers
for p phy and p beh. .
The mortality with the modelled oil drift data is considerably lower than the
estimated mortality using the field data, with only 6 and 3% within the low and
high-performance limits for the common bottlenose dolphin and the Bryde’s whale,
respectively (Table 4.7). The main reason for the large differences in the estimated
ERA Acute mortality for satellite and model oil drift data is a considerably shorter
exposure time of harmful oil in the grid cells in the modeled oil drift data (cf. Table
4.6).
For harbor seal in the PWS, the estimated mortality with ERA Acute was 195
individuals (95% CI = 28–685) resulting in an average mortality on the low side
compared to the performance boundaries (Fig. 4.13a). The reasons for the large
variation in the harbor seal results are partly due to relatively large uncertainty in the
model parameters (cf. Table 4.4) and also possibly due to a scattered distribution in
the resource dataset. The estimated mortality of or sea otters in PWS with ERA Acute
77
on the modelled oil drift data (Table 4.7). This is mainly due to considerably longer
exposure time of oil in the satellite data than in the modelled oil drift data. Compared
to the performance boundaries, all simulations resulted in mortality in the “within”
category.
4.2.2.2 Marine Mammals
The estimated mortality with ERA Acute for whales in the GOM was considerably
underestimated compared to the field assessment and the performance boundaries.
It is believed that this was primarily due to the physiological factor p phy (probability
of dying given contact with oil film on the sea surface thicker than 10 µm) firstly
being set too low for toothed and baleen whales.
Cetaceans have in general been regarded as little vulnerable towards oil spills and
the original p let factor (p phy × p beh ) was 0.1%, based on early development work for
the ERA Acute model and similar environmental risk analyses methods (e.g. FrenchMcCay 2004, 2009; Østbye et al. 2003; Spikkerud et al. 2004; Spikkerud et al. 2010).
Using the factors from previous work, the highest estimated ERA Acute mortality
for bottlenose dolphins in the Gulf of Mexico using modelled data was 228 individuals and the highest estimate using field data was 1959 individuals, considerably
underestimating the reported mortality.
During more recent development of ERA Acute, the factors were therefore
re-evaluated for toothed and baleen whales (Stephansen et al. 2018) based on
further scientific studies and preliminary reporting of high whale mortality from
the Macondo accident (DHOS) (e.g. Deepwater Horizon Natural Resource Damage
Assessment Trustees 2016). The increase in acute mortality estimated for the
common bottlenose dolphin using the refined factors is illustrated in Fig. 4 12. The
overall mean mortality increases from 1,128 (95% CI = 695–1,708) to 9,796 (95%
CI = 6,256–14,092) individuals.
Currently recommended ERA Acute parameters are presented in Table 4.4 and
are used in the results (Table 4.7), based on these calibrated vulnerability numbers
for p phy and p beh. .
The mortality with the modelled oil drift data is considerably lower than the
estimated mortality using the field data, with only 6 and 3% within the low and
high-performance limits for the common bottlenose dolphin and the Bryde’s whale,
respectively (Table 4.7). The main reason for the large differences in the estimated
ERA Acute mortality for satellite and model oil drift data is a considerably shorter
exposure time of harmful oil in the grid cells in the modeled oil drift data (cf. Table
4.6).
For harbor seal in the PWS, the estimated mortality with ERA Acute was 195
individuals (95% CI = 28–685) resulting in an average mortality on the low side
compared to the performance boundaries (Fig. 4.13a). The reasons for the large
variation in the harbor seal results are partly due to relatively large uncertainty in the
model parameters (cf. Table 4.4) and also possibly due to a scattered distribution in
the resource dataset. The estimated mortality of or sea otters in PWS with ERA Acute