View source for NL-15-029, Summary of Analysis Results Based on Corrected Format 2 Data Files Dated March 2015

{{Adams
| number = ML15252A244
| issue date = 04/06/2015
| title = Summary of Analysis Results Based on Corrected Format 2 Data Files Dated March 2015
| author name = 
| author affiliation = AKRF, Inc
| addressee name = 
| addressee affiliation = Indian Point Energy Center, NRC/NRR/DLR/RPB2
| docket = 05000247, 05000286
| license number = DPR-026, DPR-064
| contact person = 
| case reference number = NL-15-029
| document type = Report, Technical
| page count = 18
}}

=Text=
{{#Wiki_filter:Summary of Analysis Results Based on Corrected Format 2 Data Files Dated March 2015 Update of Aquatic Impact Analyses Presented in NRCs FSEIS (December 2010)
Regarding Potential Impacts of Operation of Indian Point Units 2 and 3 Prepared for:
INDIAN POINT ENERGY CENTER 450 Broadway, Suite 1 Buchanan, NY 10511 Prepared by AKRF, Inc.
7250 Parkway Drive, Suite 210 Hanover, MD 21076 April 2015

Table 3. Competing Models Used To Characterize the Standardized River Segment 4 FSS Population Trends of YOY Fish Density Using a 3-Year Moving Average (updated FSEIS Table I-9).
Species                  Linear Regression                              Segmented Regression MSE        Slope  Std Err p-value     MSE            Slope 1          Join          Slope 2 of Slope                                            Point Estimate                      Lower      Upper               Lower     Upper 95% CL 95% CL                  95% CL 95% CL Alewife                0.875      -0.055  0.026    0.047     0.863     -2.189      3.276     1989     -0.132    -0.010 American Shad          0.271      -0.117  0.014    0.000     0.249     -2.156      0.780     1988     -0.143    -0.078 Atlantic Tomcod        0.666      -0.082  0.023    0.001     0.410     -1.081     -0.239     1991     -0.080     0.023 Bay Anchovy            0.601      -0.088  0.021    0.000     0.527     -1.456      2.814     1989     -0.158    -0.063 Blueback Herring       0.877      -0.054  0.026    0.048     0.083     -5.262     -3.565     1988     -0.027     0.010 Bluefish               0.925      -0.046  0.027    0.099     0.587     0.080       1.506     1990     -0.160    -0.045 Hogchoker              0.434      -0.104  0.018    0.000     0.275     -0.271     -0.127     2001     -0.070     0.138 Rainbow Smelt          0.623      -0.086  0.022    0.001     0.535     -0.193      0.534     1992     -0.188    -0.060 Striped Bass           0.745      -0.073  0.024    0.006                         Failed to Converge Weakfish               0.810      -0.064  0.025    0.017                         Failed to Converge White Catfish          0.941      -0.042  0.027    0.128     0.892     -0.144      0.363     1995     -0.206     0.007 White Perch            0.838      -0.060  0.025    0.026     0.656     -0.543     -0.023     1995     -0.062     0.104 Table 4. River Segment 4 Assessment of the Level of Potential Negative Impact Based on the Standardized FSS Density Using a 3-Year Moving Average (updated FSEIS Table I-10).
Species        Best Fit    Slope       Slope 1        Slope 2       Final from          from          from       Decision Linear     Segmented      Segmented Regression   Regression     Regression Alewife                 SR                      S1=0          S2<0           4 American Shad           SR                      S1=0          S2<0           4 Atlantic Tomcod         SR                      S1<0          S2=0           4 Bay Anchovy             SR                      S1=0          S2<0           4 Blueback Herring        SR                      S1<0          S2=0           4 Bluefish                SR                      S1>0          S2<0           4 Hogchoker               SR                      S1<0          S2=0           4 Rainbow Smelt           SR                      S1=0          S2<0           4 Striped Bass            LR         S<0                                       4 Weakfish                LR         S<0                                       4 White Catfish           SR                      S1=0          S2=0           1 White Perch             SR                      S1<0          S2=0           4 2

Table 5. Competing Models Used To Characterize the Standardized River Segment 4 BSS Population Trends of YOY Fish Density Using a 3-Year Moving Average (updated FSEIS Table I-12).
Species                  Linear Regression                             Segmented Regression MSE        Slope  Std Err p-value     MSE            Slope 1         Join          Slope 2 of Slope                                           Point Estimate                      Lower     Upper               Lower     Upper 95% CL 95% CL                 95% CL 95% CL Alewife                0.725      0.075   0.024    0.004     0.698     -0.088     0.120     2002     -0.007     0.376 American Shad          0.215      -0.121  0.013    0.000     0.234     -0.151    -0.066     2005     -0.404     0.077 Bay Anchovy            0.844      0.059   0.025    0.029                        Failed to Converge Blueback Herring       0.726      -0.075  0.024    0.004     0.665     -0.154     0.369     1994     -0.211    -0.043 Bluefish               1.034      0.013   0.028    0.646     0.915     -0.355     0.083     1997     -0.014     0.224 Hogchoker              0.776      0.069   0.024    0.010     0.331     -0.251    -0.023     1998     0.152      0.310 Spottail Shiner        0.993      -0.030  0.028    0.290     0.930     -0.530     2.306     1989     -0.123     0.012 Striped Bass           1.023      0.019   0.028    0.502     0.447     0.110      0.343     1999     -0.292    -0.085 White Perch            1.039      -0.009  0.028    0.753     0.906     -1.099     0.153     1991     -0.042     0.111 Table 6. River Segment 4 Assessment of the Level of Potential Negative Impact Based on the Standardized BSS Density Using a 3-Year Moving Average (updated FSEIS Table I-13).
Species        Best Fit    Slope       Slope 1       Slope 2       Final from          from         from       Decision Linear     Segmented     Segmented Regression   Regression    Regression Alewife                 SR                      S1=0         S2=0           1 American Shad           LR         S<0                                      4 Bay Anchovy             LR         S>0                                      1 Blueback Herring        SR                      S1=0         S2<0           4 Bluefish                SR                      S1=0         S2=0           1 Hogchoker               SR                      S1<0         S2>0           4 Spottail Shiner         SR                      S1=0         S2=0           1 Striped Bass            SR                      S1>0         S2<0           4 White Perch             SR                      S1=0         S2=0           1 3

Table 7. Competing Models Used To Characterize the Standardized River Segment 4 LRS Population Trends of YOY Atlantic Tomcod Density Using a 3-Year Moving Average (updated FSEIS Table I-15).
Species                 Linear Regression                             Segmented Regression MSE        Slope  Std Err p-value     MSE            Slope 1       Join          Slope 2 of Slope                                        Point Estimate                      Lower     Upper             Lower     Upper 95% CL 95% CL               95% CL 95% CL Atlantic Tomcod       1.031      -0.015  0.028    0.604     0.480     -2.709    -0.671   1989     -0.007     0.090 Table 8. River Segment 4 Assessment of the Level of Potential Negative Impact Based on the Standardized LRS Atlantic Tomcod YOY Density Using a 3-Year Moving Average (updated FSEIS Table I-16).
Species        Best Fit    Slope       Slope 1       Slope 2     Final from          from         from     Decision Linear     Segmented     Segmented Regression   Regression    Regression Atlantic Tomcod        SR                     S1<0          S2=0         4 4

Table 9. Competing Models Used To Characterize the Standardized River Segment 4, FSS Population Trends of YOY Fish CPUE (updated FSEIS Table I-19).
Species                    Linear Regression                                    Segmented Regression MSE        Slope    Std Err p-value         MSE            Slope 1          Join          Slope 2 of Slope                                               Point Estimate                           Lower      Upper               Lower     Upper 95% CL 95% CL                  95% CL 95% CL Alewife                   0.953      -0.036    0.024      0.144       0.987     -0.115      0.130     2000     -0.272     0.071 American Shad             0.776      -0.063    0.022      0.007       0.688     -0.123      0.204     1996     -0.224    -0.038 Atlantic Tomcod           0.791      -0.062    0.022      0.010       0.746     -0.492      0.060     1993     -0.090     0.059 Bay Anchovy               1.039       0.003    0.025      0.903       0.882     -0.027      0.231     1999     -0.267     0.021 Blueback Herring          0.938      -0.039    0.024      0.112       0.633     -2.448     -0.120     1987     -0.048     0.049 Bluefish                  0.861      -0.052    0.023      0.031       0.847     -0.098      0.090     2002     -0.371     0.049 Hogchoker                 0.832      -0.056    0.023      0.019       0.805     -1.988      3.262     1987     -0.125    -0.022 Rainbow Smelt             0.839      -0.055    0.023      0.022       0.837     -0.265      0.451     1992     -0.163    -0.016 Striped Bass              0.952      -0.037    0.024      0.140       0.893     -1.943      3.587     1987     -0.111    -0.003 Weakfish                  1.014      -0.020    0.025      0.430       0.967     -0.091      0.130     2000     -0.296     0.092 White Perch               0.948      -0.038    0.024      0.131       0.944     -0.318      0.065     1996     -0.091     0.127 Table10. River Segment 4 Assessment of the Level of Potential Negative Impact Based on the Standardized FSS CPUE (updated FSEIS Table I-20).
Species         Best Fit       Slope         Slope 1        Slope 2       Final from            from          from       Decision Linear       Segmented      Segmented Regression     Regression     Regression Alewife                   LR           S=0                                         1 American Shad             SR                          S1=0          S2<0           4 Atlantic Tomcod           SR                          S1=0          S2=0           1 Bay Anchovy               SR                          S1=0          S2=0           1 Blueback Herring          SR                          S1<0          S2=0           4 Bluefish                  SR                          S1=0          S2=0           1 Hogchoker                 SR                          S1=0          S2<0           4 Rainbow Smelt             SR                          S1=0          S2<0           4 Striped Bass(a)           SR                          S1=0          S2=0           1 Weakfish                  SR                          S1=0          S2=0           1 White Perch               SR                          S1=0          S2=0           1 (a)
Please see the discussion of the regression analyses for striped bass following Table 26 below.
5

Table 11. Competing Models Used To Characterize the Standardized River Segment 4 LRS Population Trends of YOY Atlantic Tomcod CPUE Using a 3-Year Moving Average (updated FSEIS Table I-22).
Species                 Linear Regression                             Segmented Regression MSE        Slope  Std Err p-value     MSE            Slope 1       Join          Slope 2 of Slope                                        Point Estimate                      Lower     Upper             Lower     Upper 95% CL 95% CL               95% CL 95% CL Atlantic Tomcod       1.012      -0.021  0.025    0.410     0.842     -1.609     0.089   1988     -0.044     0.076 Table 12. River Segment 4 Assessment of the Level of Potential Negative Impact Based 7 on the Standardized LRS Atlantic Tomcod YOY CPUE Using a 3-Year Moving Average (updated FSEIS Table I-23).
Species        Best Fit    Slope       Slope 1       Slope 2     Final from          from         from     Decision Linear     Segmented     Segmented Regression   Regression    Regression Atlantic Tomcod        SR                     S1=0          S2=0         1 6

Table 13. Assessment of Population Impacts for IP2 and IP3 River Segment 4 (updated FSEIS Table I-24).
Species                 Density               CPUE           River Segment FSS     BSS      LRS     FSS    LRS      Assessment Alewife                  4        1      N/A      1      N/A         2.0 American Shad            4        4      N/A      4      N/A         4.0 Atlantic Menhaden       N/A     N/A      N/A     N/A     N/A      Unknown Atlantic Sturgeon       N/A     N/A      N/A     N/A     N/A      Unknown Atlantic Tomcod          4      N/A       4       1       1          2.5 Bay Anchovy              4        1      N/A      1      N/A         2.0 Blueback Herring         4        4      N/A      4      N/A         4.0 Bluefish                 4        1      N/A      1      N/A         2.0 Gizzard Shad            N/A     N/A      N/A     N/A     N/A      Unknown Hogchoker                4        4      N/A      4      N/A         4.0 Rainbow Smelt            4      N/A      N/A      4      N/A         4.0 Shortnose Sturgeon      N/A     N/A      N/A     N/A     N/A      Unknown Spottail Shiner         N/A       1      N/A     N/A     N/A         1.0 Striped Bass             4        4      N/A      1      N/A         3.0 Weakfish                 4      N/A      N/A      1      N/A         2.5 White Catfish            1      N/A      N/A     N/A     N/A         1.0 White Perch              4        1      N/A      1      N/A         2.0 Blue Crab               N/A     N/A      N/A     N/A     N/A      Unknown 7

Table 14. Competing Models Used To Characterize the Standardized Riverwide FSS Population Trends of YOY Fish CPUE (updated FSEIS Table I-27).
Species                Linear Regression                             Segmented Regression MSE        Slope  Std Err p-value     MSE            Slope 1         Join          Slope 2 of Slope                                           Point Estimate                      Lower     Upper               Lower     Upper 95% CL 95% CL                 95% CL 95% CL Alewife              1.005       0.023  0.025    0.361                        Failed to Converge American Shad        0.563      -0.085  0.019    0.000     0.537     -0.362     0.597     1989     -0.152    -0.050 Atlantic Tomcod      0.623      -0.080  0.020    0.000     0.628        .          .      1986     -0.128    -0.043 Bay Anchovy          1.030       0.012  0.025    0.633     0.929     -0.012     0.114     2007     -1.362     0.421 Blueback Herring     0.750      -0.066  0.021    0.005     0.795     -0.161    -0.006     2004     -0.279     0.291 Bluefish             0.750      -0.067  0.021    0.005     0.784     -0.125     0.094     1999     -0.268     0.038 Hogchoker            0.941      -0.039  0.024    0.118     0.923     -0.224     0.014     1999     -0.120     0.213 Spottail Shiner      0.838      -0.055  0.023    0.022     0.891     -0.186     0.026     2000     -0.201     0.172 Striped Bass         0.833      -0.056  0.023    0.020     0.590     0.164      2.411     1987     -0.135    -0.042 White Perch          1.022      -0.017  0.025    0.514     1.008     -0.425     0.112     1993     -0.068     0.121 Table 15. Riverwide Assessment of the Level of Potential Negative Impact Based on the Standardized FSS CPUE (updated FSEIS Table I-28).
Species        Best Fit    Slope       Slope 1       Slope 2       Final from          from         from       Decision Linear     Segmented     Segmented Regression   Regression    Regression Alewife                 LR         S=0                                      1 American Shad           SR                      S1=0         S2<0           4 Atlantic Tomcod         LR         S<0                                      4 Bay Anchovy             SR                      S1=0         S2=0           1 Blueback Herring        LR         S<0                                      4 Bluefish                LR         S<0                                      4 Hogchoker               SR                      S1=0         S2=0           1 Spottail Shiner         LR         S<0                                      4 Striped Bass            SR                      S1>0         S2<0           4 White Perch             SR                      S1=0         S2=0           1 8

Table 16. Competing Models Used To Characterize the Standardized Riverwide BSS Population Trends of YOY Fish CPUE (updated FSEIS Table I-30).
Species                  Linear Regression                             Segmented Regression MSE        Slope  Std Err p-value     MSE            Slope 1       Join          Slope 2 of Slope                                        Point Estimate                      Lower     Upper             Lower     Upper 95% CL 95% CL               95% CL 95% CL Alewife                0.701      0.072   0.021    0.002     0.687     -0.054     0.116   2002     -0.012     0.366 American Shad          0.584      -0.083  0.019    0.000     0.562     -0.358     0.623   1990     -0.161    -0.057 Atlantic Tomcod        0.507      -0.090  0.018    0.000     0.390     -0.388    -0.103   1994     -0.092     0.036 Bay Anchovy            0.800      0.061   0.022    0.011     0.593     -0.044     0.063   2006     -0.018     0.989 Blueback Herring       1.040       0.002  0.025    0.946     1.058     -1.084     1.925   1988     -0.080     0.046 Bluefish               1.036      0.008   0.025    0.754     1.048     -0.065     0.188   1999     -0.239     0.115 Hogchoker              1.025      0.015   0.025    0.546     1.054     -0.183     0.098   1998     -0.084     0.231 Rainbow Smelt          0.975      -0.031  0.024    0.210     1.005     -0.270     0.370   1993     -0.138     0.036 Spottail Shiner        0.751      0.066   0.021    0.005     0.811     -0.014     0.155   2003     -0.190     0.291 Striped Bass           1.031      0.012   0.025    0.641     0.930     -0.019     0.125   2005     -0.722     0.232 Weakfish               0.915      -0.044  0.024    0.076     0.480        .          . 1986     -0.052     0.023 White Catfish          1.034      -0.010  0.025    0.703     1.008     -1.315     0.543   1989     -0.047     0.084 White Perch            1.030      -0.012  0.025    0.624     1.000     -0.371     0.084   1994     -0.060     0.144 Table 17. Riverwide Assessment of the Level of Potential Negative Impact Based on the BSS CPUE (updated FSEIS Table I-31).
Species        Best Fit    Slope       Slope 1       Slope 2     Final from          from         from     Decision Linear     Segmented     Segmented Regression   Regression    Regression Alewife                 SR                      S1=0         S2=0         1 American Shad           SR                      S1=0         S2<0         4 Atlantic Tomcod         SR                     S1<0          S2=0         4 Bay Anchovy             SR                      S1=0         S2=0         1 Blueback Herring        LR         S=0                                    1 Bluefish                LR         S=0                                    1 Hogchoker               LR         S=0                                    1 Rainbow Smelt           LR         S=0                                    1 Spottail Shiner         LR         S>0                                    1 Striped Bass            SR                      S1=0         S2=0         1 Weakfish                LR         S=0                                    1 White Catfish           SR                      S1=0         S2=0         1 White Perch             SR                      S1=0         S2=0         1 9

Table 18. Competing Models Used To Characterize the Standardized Riverwide LRS Population Trend of YOY Atlantic Tomcod CPUE (updated FSEIS Table I-33).
Species               Linear Regression                             Segmented Regression MSE        Slope  Std Err p-value     MSE            Slope 1       Join          Slope 2 of Slope                                        Point Estimate                      Lower     Upper             Lower     Upper 95% CL 95% CL               95% CL 95% CL Atlantic Tomcod     0.831      -0.057  0.023    0.019     0.880     -0.360     0.141   1994     -0.123     0.053 Table 19. Riverwide Assessment of the Level of Potential Negative Impact Based on the Standardized LRS CPUE of Atlantic Tomcod (updated FSEIS Table I-34).
Species        Best Fit    Slope       Slope 1       Slope 2     Final from          from         from     Decision Linear     Segmented     Segmented Regression   Regression    Regression Atlantic Tomcod        LR         S<0                                    4 10

Table 20. Competing Models Used To Characterize the Standardized Riverwide YOY Abundance Index Trends (updated FSEIS Table I-36).
Species                  Linear Regression                             Segmented Regression MSE        Slope  Std Err p-value     MSE            Slope 1       Join          Slope 2 of Slope                                        Point Estimate                      Lower     Upper             Lower     Upper 95% CL 95% CL               95% CL 95% CL Alewife                1.017      0.019   0.025    0.458     0.992     -0.727     0.258   1990     -0.021     0.127 American Shad          0.596      -0.082  0.019    0.000     0.594     -0.379     0.630   1989     -0.153    -0.046 Atlantic Tomcod        0.588      -0.083  0.019    0.000     0.561     -1.665     2.716   1987     -0.141    -0.055 Bay Anchovy            0.744      -0.067  0.021    0.004     0.786     -0.194     0.005   2000     -0.198     0.152 Blueback Herring       0.792      -0.062  0.022    0.010     0.794     -0.154    -0.021   2006     -0.300     0.581 Bluefish               1.035      0.009   0.025    0.731     0.967     -0.038     0.205   1999     -0.259     0.081 Hogchoker              0.902      -0.046  0.023    0.062     0.942     -0.165     0.017   2003     -0.219     0.299 Rainbow Smelt          0.971      -0.033  0.024    0.193     0.960     -0.286     0.480   1992     -0.142     0.015 Spottail Shiner        0.844      0.055   0.023    0.024     0.879     -0.008     0.168   2003     -0.273     0.227 Striped Bass           1.039      0.004   0.025    0.864     0.928     -0.018     0.149   2003     -0.505     0.110 Weakfish               0.647      -0.077  0.020    0.001     0.576     -0.561     0.032   1992     -0.095     0.027 White Catfish          0.833      -0.056  0.023    0.020     0.863     -0.198     0.011   2001     -0.173     0.193 White Perch            1.039      -0.003  0.025    0.906     1.093     -0.079     0.103   2003     -0.424     0.243 Table 21. Riverwide Assessment of the Level of Potential Negative Impact Based in the Abundance Index (updated FSEIS Table I-37).
Species        Best Fit    Slope       Slope 1       Slope 2     Final from          from         from     Decision Linear     Segmented     Segmented Regression   Regression    Regression Alewife                 SR                      S1=0         S2=0         1 American Shad           SR                      S1=0         S2<0         4 Atlantic Tomcod         SR                     S1=0          S2<0         4 Bay Anchovy             LR         S<0                                    4 Blueback Herring        LR         S<0                                    4 Bluefish                SR                      S1=0         S2=0         1 Hogchoker               LR         S=0                                    1 Rainbow Smelt           SR                      S1=0         S2=0         1 Spottail Shiner         LR         S>0                                    1 Striped Bass            SR                      S1=0         S2=0         1 Weakfish                SR                      S1=0         S2=0         1 White Catfish           LR         S<0                                    4 White Perch             LR         S=0                                    1 11

Table 22. Assessment of Riverwide Population Impacts (updated FSEIS Table I-38).
Species                  CPUE              Abundance    Riverwide Index      Assessment FSS      BSS      LRS Alewife                   1        1       N/A         1            1.0 American Shad             4        4       N/A         4            4.0 Atlantic Menhaden        N/A      N/A      N/A        N/A        Unknown Atlantic Sturgeon        N/A      N/A      N/A        N/A        Unknown Atlantic Tomcod           4        4        4          4            4.0 Bay Anchovy               1        1       N/A         4            2.0 Blueback Herring          4        1       N/A         4            3.0 Bluefish                  4        1       N/A         1            2.0 Gizzard Shad             N/A      N/A      N/A        N/A        Unknown Hogchoker                 1        1       N/A         1            1.0 Rainbow Smelt            N/A       1       N/A         1            1.0 Shortnose Sturgeon       N/A      N/A      N/A        N/A        Unknown Spottail Shiner           4        1       N/A         1            2.0 Striped Bass              4        1       N/A         1            2.0 Weakfish                 N/A       1       N/A         1            1.0 White Catfish            N/A       1       N/A         4            2.5 White Perch               1        1       N/A         1            1.0 Blue Crab                N/A      N/A      N/A        N/A        Unknown 12

Table 23. Weight of Evidence Results for the Population Trend Line of Evidence (updated FSEIS Table H-15).
Species           River         Riverwide       WOE                Impact Segment        Assessment      Score             Conclusion Assessment        Score Score Alewife                    2.0             1.0          1.6          Undetected Decline American Shad              4.0             4.0          4.0           Detected Decline Atlantic Menhaden       Unknown         Unknown      Unknown             Unresolved Atlantic Sturgeon       Unknown         Unknown      Unknown             Unresolved Atlantic Tomcod            2.5             4.0          3.1           Detected Decline Bay Anchovy                2.0             2.0          2.0          Undetected Decline Blueback Herring           4.0             3.0          3.6           Detected Decline Bluefish                   2.0             2.0          2.0          Undetected Decline Gizzard Shad            Unknown         Unknown      Unknown             Unresolved Hogchoker                  4.0             1.0          2.8               Variable Rainbow Smelt              4.0             1.0          2.8               Variable Shortnose Sturgeon      Unknown         Unknown      Unknown             Unresolved Spottail Shiner            1.0             2.0          1.4          Undetected Decline Striped Bass               3.0             2.0          2.6               Variable Weakfish                   2.5             1.0          1.9          Undetected Decline White Catfish              1.0             2.5          1.6          Undetected Decline White Perch                2.0             1.0          1.6          Undetected Decline Blue Crab               Unknown         Unknown      Unknown             Unresolved 13

Table 24. Parameter Values Used in the Monte Carlo Simulation (updated FSEIS Table I-46).
RIS          Survey      Linear     Slope       Error      CV of     EMR        IMR Used       Slope       plus       Mean      Density (r)    Standard     Square       Data Error of     from       (1985-the Slope Regression      1996)
Estimate Alewife                BSS         0.075      0.099      0.725       1.294    0.095     0.0020 American Shad          BSS        -0.121     -0.108      0.215       0.494     0.042    0.0005 Atlantic Tomcod        FSS        -0.082     -0.059      0.666       0.812    0.036     0.0300 Bay Anchovy            FSS        -0.088     -0.067      0.601       0.510    0.213     0.0040 Blueback Herring       BSS        -0.075     -0.051      0.726       1.034     0.095    0.0040 Bluefish               BSS         0.013      0.041      1.034       0.754    0.003     0.0005 Hogchoker              FSS        -0.104     -0.086      0.434       1.224    0.386     0.0005 Rainbow Smelt          FSS        -0.086     -0.064      0.623       1.211    0.258     0.0005 Spottail Shiner        BSS        -0.030     -0.002      0.993       1.196     0.031    0.0070 Striped Bass           BSS         0.019      0.047      1.023       0.515    0.106     0.0080 Weakfish               FSS        -0.064     -0.039      0.810       0.637    0.544     0.0005 White Catfish          FSS        -0.042     -0.016      0.941       2.161    0.114     0.0005 White Perch            BSS        -0.009      0.019      1.039       1.002    0.076     0.0320 14

Table 25. Quartiles of the Relative Difference in Cumulative Abundance and Conclusions for the Strength-of-Connection From the Monte Carlo Simulation (updated FSEIS Table I-47).
Taxa         Number             N0 = 1000             N0 = 1 x 108      Strength of of     Median      Q1    Q3    Median Q1          Q3   Connection Years                                                    Conclusion Alewife                   20       -0.07    -1.19  1.03   -0.07     -1.17   1.01    Low 27       -0.32    -1.63  1.02   -0.32     -1.69   1.04 American Shad             20        0.06     0.00  0.13    0.06     -0.01   0.13    Low 27       0.05     0.00   0.11    0.06      0.00   0.11 Atlantic Tomcod           20        0.15    -0.03  0.34    0.16     -0.03   0.35    Low 27       0.15     0.01   0.30    0.15      0.01   0.29 Bay Anchovy               20        0.29     0.13  0.44    0.29      0.13   0.44    High 27       0.27     0.15   0.39    0.27      0.15   0.39 Blueback Herring          20        0.21    -0.03  0.46    0.22     -0.02   0.46    Low 27       0.22     0.04   0.41    0.23      0.04   0.42 Bluefish                  20        0.45    -0.09  0.99    0.45     -0.09   0.98    Low 27       0.67     0.11   1.21    0.69      0.15   1.23 Hogchoker                 20        0.58     0.31  0.85    0.57      0.30   0.86    High 27       0.56     0.35   0.78    0.56      0.35   0.78 Rainbow Smelt             20        0.45     0.16  0.74    0.46      0.16   0.76    High 27       0.45     0.23   0.68    0.45      0.23   0.68 Spottail Shiner           20        0.27    -0.14  0.69    0.28     -0.14   0.70    Low 27       0.34     -0.01  0.70    0.35      0.00   0.69 Striped Bass              20        0.80     0.23  1.36    0.79      0.22   1.36    High 27       1.19     0.60   1.81    1.20      0.60   1.80 Weakfish                  20        0.74     0.44  1.06    0.75      0.43   1.06    High 27       0.76     0.50   1.01    0.76      0.51   1.01 White Catfish             20        0.41    -0.18  1.01    0.43     -0.19   1.04    Low 27       0.49     0.00   0.98    0.46     -0.03   0.98 White Perch               20        0.42    -0.09  0.93    0.42     -0.07   0.91    Low 27       0.53     0.10   0.98    0.53      0.08   0.98 15

Table 26. Impingement and Entrainment Impact Summary for Hudson River YOY RIS (updated FSEIS Table H-17).
Species               Population Trend               Strength of              Impacts of IP2 and Line of Evidence              Connection              IP3 Cooling Systems Line of Evidence               on YOY RIS Alewife                     Undetected Decline                    Low                          Small American Shad               Detected Decline                      Low                          Small Atlantic Menhaden           Unresolved                           Low(b)                        Small Atlantic Sturgeon           Unresolved                           Low(b)                        Small Atlantic Tomcod             Detected Decline                      Low                          Small Bay Anchovy                 Undetected Decline                   High                          Small Blueback Herring            Detected Decline                      Low                          Small Bluefish                    Undetected Decline                    Low                          Small Gizzard Shad                Unresolved                           Low(b)                        Small Hogchoker                   Variable                             High                        Moderate Rainbow Smelt               Variable                             High                        Moderate Shortnose Sturgeon          Unresolved                           Low(b)                        Small Spottail Shiner             Undetected Decline                    Low                          Small Striped Bass                Variable                             High                        Moderate Weakfish                    Undetected Decline                   High                          Small White Catfish               Undetected Decline                    Low                          Small White Perch                 Undetected Decline                    Low                          Small Blue Crab                   Unresolved                           Low(b)                        Small (b)
Strength of connection could not be established using Monte Carlo Simulation; therefore, strength of connection was based on the rate of entrainment and impingement.
16

Discussion The conclusions presented in the February 2014 report, as well as the preceding tables, were based upon our attempt to recreate and apply the methods used by NRC in developing the conclusions presented in the FSEIS. Applying those methods to the segmented regression analysis of the River Segment 4, FSS CPUE index of abundance for striped bass, the slope 2 estimate from the February 2014 report was considered not statistically significant, but the slope 2 estimate from the current analysis is considered significant. The segmented regression analysis conducted for the February 2014 report produced a 95%
confidence interval for slope 2 from -0.1149 to 0.0005, which was considered not significant because it included zero. Whereas the current segmented regression analysis produced a 95% confidence interval from
-0.1113 to -0.0028, which was considered significant because it did not include zero.
Given the sensitivity that the overall impact conclusions can have to such small changes in statistical results, the statistical methods on which the results were based warranted a careful review. Statistical properties of the FSEIS method for determining which trend slopes are significant are discussed below along with a discussion of the use of the results of the linear regression for the striped bass data.
On page I-9 of Appendix I of the FSEIS, NRC describes its method for determining which trend slopes it considered to be significant, and how it selected a regression model on which to base trend scores:
The statistics displayed in the second table included the mean squared error (MSE) for each model; the estimate of the linear slope and associated 95 percent confidence interval; the p-value associated with the significance test of the null hypothesis that the slope (S) associated with the simple linear model equals zero; the estimated 95 percent confidence interval (CI) of the two slopes from the segmented regression (Slope 1=S1 and Slope 2=S2); and the estimated join point. For the segmented regression, slopes were defined as significant if the CI did not include zero.
The best-fit model (defined as the model with the smaller MSE) was then characterized in a third table, based on the general trend depicted by the direction of the estimated slopes. If the slope was significantly different from 0, the trend was represented by either the statement S > 0 for a positive slope or S < 0 for a negative slope. If the slope was not significant, the statement depicting the lack of a trend was S = 0.
For the segmented regressions, the FSEIS test of the null hypothesis that a slope is equal to zero (based on the confidence interval including zero) is based on the standard error of the slope estimate and is equivalent to a t-test of the significance of the slope estimate. That test is a conditional test addressing the question: after all other parameters in the model are estimated, does the inclusion of the slope estimate in the model significantly improve the fit of the model to the data? The result of the test depends on the estimates of all of the other parameters as well as the estimate of the slope being tested. Accordingly, the reliability of the other parameter estimates may affect the result of the significance test for the slope. For the striped bass, Region 4, FSS, CPUE example, NRCs method could lead to the conclusion that slope 2 was significant, although none of the other parameter estimates would have been considered statistically significant:
17

Parameter              Lower 95%               Upper 95%
Confidence Limit       Confidence Limit Join Point                1982                   1992 Slope 1                -1.9426                 3.5866 Slope 2                -0.1113                -0.0028 In addition, the FSEIS test of significant slopes does not necessarily provide an error rate of 5%, as suggested by a test based on 95% confidence intervals. As noted above, the test is equivalent to a t-test of the significance of the slope estimate. For each test based on the 95% confidence interval, the probability of falsely rejecting the null hypothesis when it is true is 5%. That is the case if one test is performed.
However, if two tests are performed (e.g. the test for slope 1 and the test for slope 2) then the probability of falsely rejecting the null hypothesis in at least one test, when both slopes in fact are zero, is greater than 0.05. For example, the probability of correctly failing to reject the null hypothesis in two independent tests when the null hypothesis is true (e.g., when both slopes are zero) is 0.95 x 0.95. In this case the probability of falsely rejecting the null hypothesis is 1 - (0.95 x 0.95) = 0.0975, almost twice the intended error probability of 0.05.
For regressions with multiple parameters being estimated simultaneously, the two issues discussed above can be avoided by testing the significance of all parameter estimates simultaneously. By doing so, the reliability of every parameter estimate is taken into account. In addition, the probability of falsely rejecting the null hypothesis is controlled and not affected by the number of parameters in the model. The standard F-test for the overall significance of a regression model can be used for this purpose. For this test, the null hypothesis is that all slopes in the model are equal to zero. The alternative hypothesis is that at least one parameter is not equal to zero (Draper, N.R. and H. Smith, 1966. Applied Regression Analysis.). If the regression is not statistically significant based on the standard F-test, then the null hypothesis that all parameters are zero cannot be rejected.
For these reasons, the standard F-test for a regression model is the appropriate test for the null hypothesis that both slopes in the segmented regression are equal to zero. The segmented regression has three parameters in addition to the intercept: two slopes and the join point. By labeling the first year in the time series as year zero, the join point parameter is equal to zero when there is no change in slope over the years being analyzed. Therefore the null hypothesis of all parameters being equal to zero would be: slope 1
= 0, slope 2 = 0 and join point = 0 (i.e., no change in slope). If the null hypothesis is rejected by the F-test, then the FSEIS conditional method could be applied to identify which particular slope estimates are significantly non-zero.
For the striped bass, River Segment 4, FSS, CPUE example, the F-statistic for the segmented regression is 2.04 (with 3 and 23 degrees of freedom). The probability of observing an F-statistic that large under the null hypothesis is 0.14. Because 0.14 exceeds the acceptable error rate for the test of 0.05, the null hypothesis that both slopes are equal to zero cannot be rejected. Because the null hypothesis that all slopes are equal to zero cannot be rejected, it is not necessary or appropriate to apply the FSEIS conditional test for the significance of individual slopes in this case. This more robust statistical result that neither slope 1 nor slope 2 is significantly different from zero is consistent with the results presented in the February 2014 report.
18}}