| dc.identifier.uri | http://hdl.handle.net/11401/76207 | |
| dc.description.sponsorship | This work is sponsored by the Stony Brook University Graduate School in compliance with the requirements for completion of degree. | en_US |
| dc.format | Monograph | |
| dc.format.medium | Electronic Resource | en_US |
| dc.language.iso | en_US | |
| dc.publisher | The Graduate School, Stony Brook University: Stony Brook, NY. | |
| dc.type | Thesis | |
| dcterms.abstract | Storm surge from extratropical cyclones can lead to significant coastal flooding, particularly along coastlines adjacent to wide and shallow continental shelves. In warming global climate simulations, studies have shown systematic regional changes to cyclones, which generate surges. However, there have been no formal studies that link regional cyclone changes to changes in the occurrence and intensity of surges along the densely populated New York/New Jersey coastline. A multi-linear regression (MLR) approach is developed to predict 3-hourly surge during the cool season months (Oct. 1-March 31) using climate model data. At the three stations the approach is applied to (The Battery in New York, Atlantic City, New Jersey, and Montauk Point, New York), the MLR explains > 60.0 % of the observed surge variance in 3-h surge data. Predictions of storm maximum surge that meet or exceed the 95th percentile of storm maximum surges have a mean absolute error between 0.30 – 0.40 m and a mean error around zero. Using the same forecasted surface winds and pressures from the North American Mesoscale (NAM) model between October-March 2010 to 2014, surge predictions at The Battery are compared to raw output from a numerical hydrodynamic model's (SIT-NYHOPS) predictions. The accuracy of surge predictions between the SIT-NYHOPS and the MLR at The Battery are similar for predictions that meet or exceed the 95th percentile of storm maximum surge heights at 0-24 h leads. The MLR is applied to an ensemble of six global climate models part of the 5th Generation of the Coupled Model Intercomparison Project (CMIP5) to study future surge changes at The Battery, New York. The predicted surges using climate models have realistic amplitudes, but the seasonal frequency of impactful surge events (≥0.61 m) is underpredicted by 2-5 events. While there are large interseasonal variations in storm surge, the majority of the models demonstrate no trends or intensity changes greater than natural or historically modeled surge variability. Only one GCM (CCSM4) predicts a noticeable shift toward more intense surge events in 2054-2079 compared to 1979-2004, which is explained with a northward shift in storm track. However, the effects of a regional sea level rise (SLR) scenario are much larger than any future modeled surge changes. Approximately 10 times more moderate coastal floods (≥2.44 m above MLLW) are predicted in 2069-2079 compared to 2009-2019 with the addition of a regional sea level rise scenario. The intensity, spatial, and frequency distributions of classified tracks are related to surge events at The Battery to help explain the mechanisms that affect surge climatology. Surface cyclone tracks were automatically tracked in mean sea level pressure fields using the Hodges cyclone-tracking algorithm and were matched in time to modeled and observed surge time series. Tracks were automatically classified as either Miller Type A or Miller Type B following a set of rules. Both Miller Type A and Miller Type B cyclone tracks are shown to generate the majority (60/75) of observed impactful surge events at The Battery between 1979-2004 Nov.-March. The interseasonal frequency of Miller Type A tracks is shown to correlate positively with the interseasonal variations in the frequency of impactful surges in both observed (1979-2004) and modeled data (1979-2079). Between 2054-2079 and 1979-2004, there are no modeled changes that are consistent between models in the spatial, intensity, and frequency distribution of surge-generating cyclone tracks. | |
| dcterms.abstract | Storm surge from extratropical cyclones can lead to significant coastal flooding, particularly along coastlines adjacent to wide and shallow continental shelves. In warming global climate simulations, studies have shown systematic regional changes to cyclones, which generate surges. However, there have been no formal studies that link regional cyclone changes to changes in the occurrence and intensity of surges along the densely populated New York/New Jersey coastline. A multi-linear regression (MLR) approach is developed to predict 3-hourly surge during the cool season months (Oct. 1-March 31) using climate model data. At the three stations the approach is applied to (The Battery in New York, Atlantic City, New Jersey, and Montauk Point, New York), the MLR explains > 60.0 % of the observed surge variance in 3-h surge data. Predictions of storm maximum surge that meet or exceed the 95th percentile of storm maximum surges have a mean absolute error between 0.30 – 0.40 m and a mean error around zero. Using the same forecasted surface winds and pressures from the North American Mesoscale (NAM) model between October-March 2010 to 2014, surge predictions at The Battery are compared to raw output from a numerical hydrodynamic model's (SIT-NYHOPS) predictions. The accuracy of surge predictions between the SIT-NYHOPS and the MLR at The Battery are similar for predictions that meet or exceed the 95th percentile of storm maximum surge heights at 0-24 h leads. The MLR is applied to an ensemble of six global climate models part of the 5th Generation of the Coupled Model Intercomparison Project (CMIP5) to study future surge changes at The Battery, New York. The predicted surges using climate models have realistic amplitudes, but the seasonal frequency of impactful surge events (≥0.61 m) is underpredicted by 2-5 events. While there are large interseasonal variations in storm surge, the majority of the models demonstrate no trends or intensity changes greater than natural or historically modeled surge variability. Only one GCM (CCSM4) predicts a noticeable shift toward more intense surge events in 2054-2079 compared to 1979-2004, which is explained with a northward shift in storm track. However, the effects of a regional sea level rise (SLR) scenario are much larger than any future modeled surge changes. Approximately 10 times more moderate coastal floods (≥2.44 m above MLLW) are predicted in 2069-2079 compared to 2009-2019 with the addition of a regional sea level rise scenario. The intensity, spatial, and frequency distributions of classified tracks are related to surge events at The Battery to help explain the mechanisms that affect surge climatology. Surface cyclone tracks were automatically tracked in mean sea level pressure fields using the Hodges cyclone-tracking algorithm and were matched in time to modeled and observed surge time series. Tracks were automatically classified as either Miller Type A or Miller Type B following a set of rules. Both Miller Type A and Miller Type B cyclone tracks are shown to generate the majority (60/75) of observed impactful surge events at The Battery between 1979-2004 Nov.-March. The interseasonal frequency of Miller Type A tracks is shown to correlate positively with the interseasonal variations in the frequency of impactful surges in both observed (1979-2004) and modeled data (1979-2079). Between 2054-2079 and 1979-2004, there are no modeled changes that are consistent between models in the spatial, intensity, and frequency distribution of surge-generating cyclone tracks. | |
| dcterms.available | 2017-09-20T16:49:39Z | |
| dcterms.contributor | Hameed, Sultan | en_US |
| dcterms.contributor | Colle, Brian A. | en_US |
| dcterms.contributor | Bowman, Malcolm. | en_US |
| dcterms.creator | Roberts, Keith J. | |
| dcterms.dateAccepted | 2017-09-20T16:49:39Z | |
| dcterms.dateSubmitted | 2017-09-20T16:49:39Z | |
| dcterms.description | Department of Marine and Atmospheric Science. | en_US |
| dcterms.extent | 139 pg. | en_US |
| dcterms.format | Monograph | |
| dcterms.format | Application/PDF | en_US |
| dcterms.identifier | http://hdl.handle.net/11401/76207 | |
| dcterms.issued | 2015-12-01 | |
| dcterms.language | en_US | |
| dcterms.provenance | Made available in DSpace on 2017-09-20T16:49:39Z (GMT). No. of bitstreams: 1
Roberts_grad.sunysb_0771M_12275.pdf: 15880372 bytes, checksum: 7f930274e882533bd0bd81d820457d45 (MD5)
Previous issue date: 1 | en |
| dcterms.publisher | The Graduate School, Stony Brook University: Stony Brook, NY. | |
| dcterms.subject | Regression Analysis, Storm surge | |
| dcterms.subject | Atmospheric sciences | |
| dcterms.title | An Application of Regression for Storm Surge Prediction along the New York/New Jersey Coast in Climate Models | |
| dcterms.type | Thesis | |
