Prepared in cooperation with the Bureau of Land Management
Forecasts of Polar Bear (Ursus maritimus) Land Use in the Southern Beaufort and Chukchi Seas, 2040–65
Open-File Report 2023–1048
U.S. Department of the Interior U.S. Geological Survey
Forecasts of Polar Bear (Ursus maritimus) Land Use in the Southern Beaufort and Chukchi Seas, 2040–65 By Karyn D. Rode, David C. Douglas, Todd C. Atwood, and Ryan R. Wilson
Prepared in cooperation with the Bureau of Land Management
Open-File Report 2023–1048
U.S. Department of the Interior U.S. Geological Survey
U.S. Geological Survey, Reston, Virginia: 2023
For more information on the USGS—the Federal source for science about the Earth, its natural and living resources, natural hazards, and the environment—visit https://www.usgs.gov or call 1–888–392–8545. For an overview of USGS information products, including maps, imagery, and publications, visit https://store.usgs.gov/ or contact the store at 1–888–275–8747. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Although this information product, for the most part, is in the public domain, it also may contain copyrighted materials as noted in the text. Permission to reproduce copyrighted items must be secured from the copyright owner. Suggested citation: Rode, K.D., Douglas, D.C., Atwood, T.C., and Wilson, R.R., Forecasts of polar bear (Ursus maritimus) land use in the southern Beaufort and Chukchi Seas, 2040–65: U.S. Geological Survey Open-File Report 2023–1048, 7 p., https://doi.org/10.3133/ofr20231048. Associated data for this publication: Rode, K. D., Douglas, D. C., Atwood, T. C., Durner, G. M., Wilson, R. R., Bromaghin, J. F., Pagano, A. M. and Simac, K. S., 2022, Polar bear Continuous Time-Correlated Random Walk (CTCRW) location data derived from satellite location data, Chukchi and Beaufort Seas, July-November 1985-2017: U.S. Geological Survey data release, https://doi.org/10.5066/P9XEOBWV. ISSN 2331-1258 (online)
iii
Acknowledgments Our ability to examine long-term patterns in polar bear land use in the southern Beaufort and Chukchi Seas is made possible by the dedication of biologists and pilots who, over many years, captured and collared polar bears in this region. These persons include but are not limited to E. Regehr, M. St. Martin, S. Amstrup, C. Gardner, S. Arthur, G. Durner, S. Knick, G. Garner, G. York, K. Simac, L. Peacock, and T. Donnelly, and pilots from Maritime Helicopters, Prism Helicopters, and Soloy Helicopters LLC. We thank the communities of Utqiağvik, Kaktovik, Point Hope, and Kivalina and the North Slope Borough for accommodating our field efforts. We acknowledge the World Climate Research Programme's Working Group on Coupled Modelling, which is responsible for the Coupled Model Intercomparison Project (CMIP). For CMIP, the U.S. Department of Energy Program for Climate Model Diagnosis and Intercomparison provided coordinating support and led development of software infrastructure in partnership with the Global Organization for Earth System Science Portals. We thank J. Bromaghin, E. Andersen, and C. Perham for providing reviews that helped improve the report.
v
Contents Acknowledgments����������������������������������������������������������������������������������������������������������������������������������������iii Abstract�����������������������������������������������������������������������������������������������������������������������������������������������������������1 Introduction����������������������������������������������������������������������������������������������������������������������������������������������������1 Methods����������������������������������������������������������������������������������������������������������������������������������������������������������2 Results�������������������������������������������������������������������������������������������������������������������������������������������������������������2 Discussion�������������������������������������������������������������������������������������������������������������������������������������������������������7 References Cited�������������������������������������������������������������������������������������������������������������������������������������������7
Figures 1.
Graphs showing forecasts of the percent of polar bears summering onshore and their duration onshore in the southern Beaufort Sea and Chukchi Sea populations estimated from the slope of trend in projected sea-ice conditions by earth system models and observed relationships between changes in sea ice availability and changes in land use�����������������������������������������������������������������������������������4
Tables 1.
2. 3. 4. 5.
Estimates of the percentage of polar bears summering onshore in the southern Beaufort Sea and Chukchi Sea populations and their estimated duration onshore in the year 2065 for three greenhouse gas emission scenarios from global circulation models������������������������������������������������������������������������������������������������������������3 Estimates of the percent of polar bears summering onshore in the southern Beaufort Sea population forecasted for every 5 years from 2040 to 2065 for three greenhouse gas emission scenarios from global circulation models�����������������������5 Estimates of the duration polar bears may summer onshore in the southern Beaufort Sea population forecasted for every 5 years from 2040 to 2065 for three greenhouse gas emission scenarios from global circulation models�����������������������5 Estimates of the percentage of polar bears summering onshore in the Chukchi Sea population forecasted for every 5 years from 2040 to 2065 for three greenhouse gas emission scenarios from global circulation models���������������������������������6 Estimates of the duration polar bears may summer onshore in the Chukchi Sea population forecasted for every 5 years from 2040 to 2065 for three greenhouse gas emission scenarios from global circulation models��������������������������������������������������������6
vi
Abbreviations CMIP6
Coupled Model Intercomparison Project Phase 6
ESM
Earth System Model
CS SB SSP1-2.6 SSP2-4.5 SSP5-8.5
Chukchi Sea southern Beaufort Sea greenhouse gas emission scenario SSP 1-2.6 greenhouse gas emission scenario SSP 2-4.5 greenhouse gas emission scenario SSP 5-8.5
Forecasts of Polar Bear (Ursus maritimus) Land Use in the Southern Beaufort and Chukchi Seas, 2040–65 By Karyn D. Rode1, David C. Douglas1, Todd C. Atwood1, and Ryan R. Wilson2
Abstract This report provides analysis to extend the 2040 forecasts of polar bear (Ursus maritimus) land use for the southern Beaufort and Chukchi Sea populations presented in a recent publication (Rode and others, 2022) through the year 2065. To inform long-term polar bear management considerations, we provide point-estimate forecasts and 95-percent prediction intervals of the proportion of polar bear populations summering onshore for 21 days or more (≥) and their duration onshore every 5 years from 2040 to 2065. Because sea-ice projections based on earth system models show greater divergence with emission scenarios after 2040, we have provided forecasts for three greenhouse gas emission scenarios, SSP1-2.6, SSP2-4.5 and SSP5-8.5, compared to the two emission scenarios used in Rode and others (2022). Our forecasting methods estimated that 61–97 percent of polar bears in the southern Beaufort Sea and 80–100 percent of polar bears in the Chukchi Sea populations may summer onshore for ≥21 days by 2065. Forecasts of mean duration onshore were 105–158 days for polar bears in the southern Beaufort Sea and 111–178 days in the Chukchi Sea populations by 2065. Sea ice conditions projected to occur by 2065 could alter the current patterns of bear behavior from what has been observed over the past 30 years. As a result, these extended projections are associated with a higher degree of uncertainty than estimates through 2040, especially under the SSP5-8.5 scenario.
Introduction In November 2022, U.S. Geological Survey scientists, in collaboration with the U.S. Fish and Wildlife Service, forecasted summer land use by polar bears (Ursus maritimus) in the two populations whose range occurs in Alaska: the southern Beaufort Sea (SB) and Chukchi Sea (CS) populations (Rode and others, 2022). Location data collected from collared female polar bears in each population were used to identify the 1 U.S. Geological Survey 2 U.S. Fish and Wildlife Service
percentage of bears that summered on land and the relations between land use and sea-ice conditions between the 1980s and 2017. Forecasts of future land use were made using two approaches: (1) extrapolating the observed trends between the 1980s and 2017 and (2) using sea-ice projections from earth systems models (ESMs) in combination with observed relations between land use and sea-ice conditions. Here, we expand on this earlier work to forecast summer land use to the year 2065 in response to needs expressed by management agencies for forecasting beyond the previously-published period of 2023–40. Forecasts based on sea-ice projections (hereinafter “sea-ice-based forecast”) used sea-ice projections estimated directly from global models when forced with different greenhouse gas emission scenarios. However, these sea-ice-based forecasts resulted in values that were below the endpoints of present-day trends in observed sea-ice conditions. This tendency for global climate models to overestimate current sea-ice extent has been referred to as “faster than forecast” (Stroeve and others, 2007). To compensate for these apparent overestimates, we have applied an ad hoc adjustment as described below and in Rode and others (2022), and we hereinafter refer to them as “adjusted-sea ice-based forecasts.” For both sea-ice-based forecasts, we diverge from Rode and others (2022) by analyzing three, rather than two, emission scenarios. Rode and others (2022) forecasted summer land use using sea-ice projections based on SSP1-2.6, in which greenhouse gas emissions are promptly and aggressively mitigated, and SSP5-8.5, in which atmospheric greenhouse gas concentrations rise unabated. Nevertheless, sea-ice projections showed little to no variation with emission scenario through 2040. However, after 2040, sea-ice projections begin to diverge with emission scenario. Therefore, in extending our forecasts to 2065, we added the SSP2-4.5 scenario, which represents a middle-of-the-road scenario in which emissions and land-use pathways are not extreme (O’Neill and others, 2016).
2 Forecasts of Polar Bear Land Use, 2040–65
Methods Detailed methods used to generate the land-use forecasts are presented in Rode and others (2022). Here, rather than providing a single forecast for 2065, we provide point estimates and prediction intervals every 5 years during 2040–65 to assist in management planning. The results for 2040 presented here are identical to those in Rode and others (2022). In Rode and others (2022), forecasts were made using three methods: (1) extrapolation of the observed trend in land use over the previous three decades, (2) use of sea-ice projections from earth system models (ESMs) with observed relationships between sea-ice availability and land use, and (3) use of the linear slope in sea-ice projections from ESMs with an adjustment on the y-axis to begin the projected sea-ice trend at the value obtained from regressing the past 20 years of observational sea-ice data. Here, we used the same three methods as we applied previously but for method 3, we generated point estimate forecasts from sea-ice projections in 5-year increments. Because the trajectory of the sea-ice projections was not always linear depending on the end year of the forecast, the slope of the ESM multi-model sea-ice projection was specific to each end year and emission scenario (Rode and others, 2022, fig. S2). These longer-range forecasts in comparison to forecasting in Rode and others (2022) resulted in some of the forecasted estimates for the percentage of the population summering onshore exceeding 100 percent.
Because that is not plausible, we report the estimates capped at 100 percent. We present results for both the sea-ice-based and adjusted-sea-ice-based forecasts but focus our summary on the latter forecast in addition to the extrapolated forecast.
Results A summary of forecasted point estimates in 2065 using all three forecasting methods is provided in table 1. Figure 1 shows forecasted point estimates and prediction intervals for each land-use metric (percentage onshore and duration onshore) for the SB and CS populations every 5 years from 2040 to 2065 based on method 1 (extrapolation) and method 3 (adjusted sea-ice-based forecast). Results in figure 1 are presented in tables 2–5, accompanied by additional results obtained with forecasting method 2 (sea-ice projections that were not corrected for faster than forecast). Polar bear land-use forecasts for 2040 through 2065 differed considerably depending on the greenhouse gas emissions scenario used by ESMs to project future sea-ice conditions (fig. 1). Under the SSP1-2.6 scenario, sea-ice loss after mid-century stabilizes so polar bear land-use forecasts based on sea ice (methods 2 and 3) also stabilize. In contrast, under SSP5-8.5, forecasted land use by polar bears continues to increase as more sea ice melts.
Results 3 Table 1. Estimates of the percentage of polar bears (Ursus maritimus) summering onshore in the southern Beaufort Sea (SB) and Chukchi Sea (CS) populations and their estimated duration onshore in the year 2065 for three greenhouse gas emission scenarios from global circulation models (SSP1-2.6, 2-4.5, and 5-8.5). [Ninety-five percent prediction intervals are shown in parentheses adjacent to the point estimates. NA, not applicable]
Scenario
NA
SB Percentage onshore 75 (31–100)
CS Duration (days)
Percentage onshore
Method 1—Extrapolation 140 (96–184)
Method 2—Sea-ice-based forecast
Duration (days)
100 (48–100)
178 (108–247)
SSP1-2.6
51 (21–80)
91 (53–130)
75 (32–100)
105 (56–154)
SSP2-4.5
72 (32–100)
123 (77–168)
97 (44–100)
130 (74–186)
SSP5-8.5
80 (36–100)
135 (86–184)
100 (46–100)
135 (78–193)
SSP1-2.6
61 (26–95)
SSP2-4.5 SSP5-8.5
Method 3—Adjusted-sea-ice-based forecast 105 (64–146)
80 (34–100)
111 (60–161)
79 (35–100)
132 (84–180)
100 (48–100)
138 (80–197)
97 (43–100)
158 (104–214)
100 (53–100)
150 (87–212)
4 Forecasts of Polar Bear Land Use, 2040–65
Figure 1. Forecasts of the percent of polar bears (Ursus maritimus) summering onshore and their duration onshore in the southern Beaufort Sea and Chukchi Sea populations estimated from the slope of trend in projected sea-ice conditions by earth system models and observed relationships between changes in sea ice availability and changes in land use. Forecasts are provided in 5-year increments for three greenhouse gas emission scenarios.
Results 5 Table 2. Estimates of the percent of polar bears (Ursus maritimus) summering onshore in the southern Beaufort Sea population forecasted for every 5 years from 2040 to 2065 for three greenhouse gas emission scenarios from global circulation models (SSP 1-2.6, 2-4.5, and 5-8.5). [Ninety-five percent prediction intervals are shown in parentheses adjacent to the point estimates. NA, not applicable]
Scenario
Year 2040
2045
2050
2055
2060
2065
70 (29–100)
75 (31–100)
Method 1—Extrapolation NA
51 (20–82)
56 (22–89)
60 (24–97)
65 (27–100)
Method 2—Sea-ice-based forecast
SSP1-2.6
41 (15–67)
51 (21–80)
54 (23–85)
48 (19–76)
56 (23–86)
51 (21–80)
SSP2-4.5
52 (22–83)
53 (22–84)
58 (25–92)
60 (26–93)
64 (27–98)
72 (32–100)
SSP5-8.5
55 (23–87)
62 (27–98)
65 (28–100)
78 (35–100)
77 (34–100)
80 (36–100)
Method 3—Adjusted-sea-ice-based forecast
SSP1-2.6
50 (21–80)
55 (24–87)
59 (26–93)
60 (26–94)
62 (27–97)
61 (26–95)
SSP2-4.5
52 (22–83)
57 (25–90)
65 (28–100)
68 (30–100)
74 (32–100)
79 (35–100)
SSP5-8.5
62 (27–96)
75 (33–100)
81 (36–100)
89 (40–100)
94 (42–100)
97 (43–100)
Table 3. Estimates of the duration (in days) polar bears (Ursus maritimus) may summer onshore in the southern Beaufort Sea population forecasted for every 5 years from 2040 to 2065 for three greenhouse gas emission scenarios from global circulation models (SSP 1-2.6, 2-4.5, and 5-8.5). [Ninety-five percent prediction intervals are shown in parentheses adjacent to the point estimates. NA, not applicable]
Scenario
NA
Year 2040
2045
2050
98 (65–132)
106 (71–142)
Method 1—Extrapolation 115 (77–152)
2055
2060
2065
123 (83–163)
131 (90–173)
140 (96–184)
Method 2—Sea-ice-based forecast
SSP1-2.6
77 (41–113)
91 (53–130)
96 (57–136)
87 (50–125)
97 (58–137)
91 (53–130)
SSP2-4.5
94 (55–133)
95 (56–134)
103 (62–144)
105 (63–146)
109 (67–151)
123 (77–168)
SSP5-8.5
98 (59–138)
109 (67–151)
112 (69–155)
131 (84–179)
130 (83–178)
135 (86–184)
107 (65–149)
105 (64–146)
SSP1-2.6
Method 3—Adjusted-sea-ice-based forecast
90 (52–128)
98 (58–137)
103 (62–144)
105 (63–146)
SSP2-4.5
94 (55–131)
100 (60–140)
111 (68–154)
116 (72–160)
122 (77–168)
132 (84–180)
SSP5-8.5
108 (66–149)
127 (80–174)
136 (87–186)
147 (95–200)
155 (100–209)
158 (103–214)
6 Forecasts of Polar Bear Land Use, 2040–65 Table 4. Estimates of the percentage of polar bears (Ursus maritimus) summering onshore in the Chukchi Sea population forecasted for every 5 years from 2040 to 2065 for three greenhouse gas emission scenarios from global circulation models (SSP 1-2.6, 2-4.5, and 5-8.5). [Ninety-five percent prediction intervals are shown in parentheses adjacent to the point estimates. NA, not applicable]
Scenario
NA
Year 2040 79 (31–100)
2045
2050 Method 1—Extrapolation
86 (35–100)
92 (38–100)
2055
2060
2065
99 (41–100)
100 (45–100)
100 (48–100)
Method 2—Sea-ice-based forecast
SSP1-2.6
69 (27–100)
73 (30–100)
75 (31–100)
84 (36–100)
81 (35–100)
75 (31–100)
SSP2-4.5
72 (29–100)
77 (33–100)
80 (34–100)
87 (39–100)
90 (40–100)
97 (44–100)
SSP5-8.5
82 (36–100)
88 (39–100)
86 (38–100)
200 (46–100)
99 (46–100)
100 (46–100)
SSP1-2.6
79 (34–100)
81 (35–100)
84 (37–100)
80 (34–100)
Method 3—Adjusted-sea-ice-based forecast 85 (37–100)
83 (36–100)
SSP2-4.5
82 (36–100)
84 (37–100)
90 (40–100)
94 (43–100)
98 (45–100)
100 (48–100)
SSP5-8.5
87 (39–100)
99 (45–100)
100 (47–100)
100 (50–100)
100 (52–100)
100 (53–100)
Table 5. Estimates of the duration (in days) polar bears (Ursus maritimus) may summer onshore in the Chukchi Sea population forecasted for every 5 years from 2040 to 2065 for three greenhouse gas emission scenarios from global circulation models (SSP 1-2.6, 2-4.5, and 5-8.5). [Ninety-five percent prediction intervals are shown in parentheses adjacent to the point estimates. NA, not applicable]
Scenario
NA
Year 2040 126 (74–179)
2045
2050 Method 1—Extrapolation
137 (81–192)
147 (88–206)
2055
2060
2065
157 (95–220)
168 (102–233)
178 (108–247)
112 (61–162)
105 (56–154)
Method 2—Sea-ice-based forecast
SSP1-2.6
98 (51–145)
103 (54–151)
105 (56–154)
114 (63–165)
SSP2-4.5
101 (53–149)
107 (58–157)
110 (60–161)
119 (67–172)
122 (69–176)
130 (74–186)
SSP5-8.5
113 (62–165)
119 (67–172)
118 (66–170)
134 (77–191)
133 (76–190)
135 (78–193)
SSP1-2.6
110 (60–160)
112 (61–163)
116 (64–168)
115 (64–167)
111 (60–161)
Method 3—Adjusted-sea-ice-based forecast
114 (63–166)
SSP2-4.5
113 (62–164)
115 (63–166)
122 (69–176)
127 (72–183)
132 (75–189)
138 (80–197)
SSP5-8.5
119 (66–172)
133 (76–190)
137 (79–195)
142 (82–202)
147 (85–208)
150 (87–212)
References Cited 7
Discussion Rode and others (2022) reported that over the past three decades, the percentage of polar bears summering onshore increased from 5 to 30 percent in the southern Beaufort Sea (SB) and 10 to 50 percent in the Chukchi Sea (CS). Duration onshore for both populations increased by more than 30 days to 60–70 days between the 1980s and 2020. Based on the extrapolation and adjusted-sea-ice-based forecasts, by 2040, an estimated 50–62 percent of SB and 79–87 percent of CS polar bears were forecasted to respectively spend 90–108 and 110–126 days onshore during the summer months. The forecasts presented here, which extend to 2065, estimate that if patterns observed over the past 30 years continue (that is, the forecast based on extrapolations in tables 2 and 4), 75 percent of polar bears in the SB and 100 percent of polar bears in the CS would summer onshore. Estimates of the percent of bears onshore varied widely with greenhouse gas emission scenario for the adjusted sea-ice-based forecasts (tables 2 and 4) from as low as 61 percent in the SB under SSP1-2.6 to as high as 97 percent under SSP5-8.5. Extrapolation of observed linear trends estimated that polar bears would spend 140 and 178 days summering onshore in the SB and CS, respectively, by 2065. In the SB, forecasted time frame onshore using adjusted sea-ice projections estimated that polar bears would spend 105–158 days onshore by 2065 depending on emission scenario. In the CS, the extrapolated estimate of duration onshore was higher than the range using adjusted sea-ice projections estimated across emission scenarios (111–150 days). Forecasting polar bear behavior 45 years into the future based on correlations with past observations inherently entails high uncertainty. Rode and others (2022, p. 17) postulated that “the generalized response by polar bears to increasingly use land habitats in response to sea ice loss is likely to continue beyond 2040 and be affected by the level of greenhouse gas emissions” but chose not to estimate land-use metrics beyond 2040 because the observational relations underlying the forecasts could change. For example, many of the CMIP6 ESMs used in this study project that by 2065 sea ice will melt entirely in late summer, especially under the SSP5-8.5 emissions scenario (Douglas and Atwood, 2022). Polar bears
choosing to remain on sea ice throughout the summer would need to migrate with the retreating ice pack into the high latitudes of northwestern Canada and northern Greenland (where remnants of sea ice will likely persist the longest) and ultimately come ashore in those areas if the ice melted completely. Such novel environmental conditions are not explicitly represented in the observed responses of polar bears to past sea-ice changes that underlie the forecasts generated here, so the resulting land-use estimates possess a greater degree of uncertainty, especially under the high emissions scenario.
References Cited Douglas, D.C., and Atwood, T.C., 2022, Comparisons of Coupled Model Intercomparison Project Phase 5 (CMIP5) and Coupled Model Intercomparison Project Phase 6 (CMIP6) sea-ice projections in polar bear (Ursus maritimus) ecoregions during the 21st century: U.S. Geological Survey Open-File Report 2022–1062, 27 p. [Also available at https://doi.org/10.3133/ofr20221062.] O’Neill, B.C., Tebaldi, C., Van Vuuren, D.P., Eyring, V., Friedlingstein, P., Hurtt, G., Knutti, R., Kriegler, E., Lamarque, J-F., Lowe, J., Meehl, G.A., Moss, R., Riahi, K., and Sanderson, B.M., 2016, The Scenario Model Intercomparison Project (ScenarioMIP) for CMIP6: Geoscientific Model Development, v. 9, no. 9, p. 3461–3482 [Also available at https://doi.org/10.5194/gmd-9-461-2016.] Rode, K.D., Douglas, D.C., Atwood, T.C., Durner, G.M., Wilson, R.R., and Pagano, A.M., 2022, Observed and forecasted changes in land use by polar bears in the Beaufort and Chukchi Seas, 1985–2040: Global Ecology and Conservation, v. 40, p. e02319 [Also available at https://doi.org/10.1016/j.gecco.2022.e02319.] Stroeve, J., Holland, M.M., Meier, W., Scambos, T., and Serreze, M., 2007, Arctic sea ice decline—Faster than forecast: Geophysical Research Letters, v. 34, no. 9, p. L09501 [Also available at https://doi.org/10.1029/ 2007GL029703.]
For more information about the research in this report, contact Director, Alaska Science Center U.S. Geological Survey 4210 University Drive Anchorage, Alaska 99508 https://www.usgs.gov/centers/alaska-science-center Manuscript approved on June 12, 2023 Publishing support provided by the U.S. Geological Survey Science Publishing Network, Tacoma Publishing Service Center
Rode and others—Forecasts of Polar Bear Land Use, 2040–65—OFR 2023–1048
ISSN 2331-1258 (online) https://doi.org/10.3133/ofr20231048