Introduction
The purpose of this paper is to accompany the 3-D model for sea level rise affecting New York City. It will begin with a background of the forces affecting future sea level rise, how much sea level rise would affect New York City, and how such a 3-D model can be useful is visualizing such effects. This will give way to an explanation of exactly what Geographic Information System (GIS) techniques were used to create the 3-D model. Finally, the results of the model will be laid out for a better understanding of what the model itself is depicting.
Climate change is a major concern for a myriad of reasons. Rising global temperatures will have either a direct or indirect effect on nearly every facet of life that we as humans often don’t think twice about. Rising surface temperatures will yield an increase in hurricanes, landslides, wildfires, floods, tidal waves, melting permafrost, severe drought, malnutrition, human migration, strain on infrastructure, and not least of all the melting of the polar ice caps which will in turn incrementally raise the global sea level.
But sea level rise is not a new phenomenon. For most of the twentieth century, sea level rise remained under 2mm/year. This is correlated with global temperatures rising only about 1 degree Celsius over the same period. This stands in stark contrast to more recent rates of sea level rise, which rose 3.6mm/year from 2006 to 2015. Left unchecked in a worst-case scenario the global sea level can rise nearly seven feet by 2100 and 13 feet by 2150 over the level recorded in 2000. With nearly 30% of the population of the United States living in coastal areas that are vulnerable to sea level rise, this can be disastrous.
New York City, the area of focus for this project, is one of those areas. Already flooded by coastal waters several times from hurricanes over the previous 20 years, the city’s vulnerability to an increase in water levels is apparent considering the $19 billion in damages from Hurricane Sandy alone. But in the case of global sea level rise, the is no natural drainage for the water – it’s all here to stay. With that said, the city has not built confidence in its resiliency to future water-based emergencies, least of all a drastic rise in global sea levels. Current plans include restructuring parks (cutting down about 1,000 mature trees), artificially elevating some areas up to nine feet, and creating floodwalls and floodgates.
Exacerbating the impact of sea level rise in New York City is the city’s large population that resides there. Currently, 1.3 million people reside in the current floodplain. This number is expected to increase to 2.2 million by 2100; considering that the floodplain will likely also cover a larger area, this number leans on the conservative side. More than half of the current population that lives within the current floodplain reside in areas with a median yearly income of less than $75,000, which marks the “low income” threshold for a family of three. Being low-income and largely under-resourced minority groups living in the floodplain, these communities experience greater challenges evacuating, increased debt, a lack of adequate medical care, and overall, an increased risk of death.
Methods
Two programs were used in the processing of data and creating of this 3-D modeling project. The first is Esri’s (Environmental Systems Research Institute) ArcMap program, and the second is Esri’s ArcScene program. ArcMap has been the powerhouse premier GIS application for decades, being able to combine maps and data for spatial analysis. ArcScene is a spinoff program designed for the creation of 3-D models, exactly applicable to the needs of this project.
Data for this project were obtained from various sources. Elevation data was found on the New York City Open Data site, yielding a LiDAR-based one-foot resolution digital elevation model. LiDAR means Light Detection and Ranging, which is similar to Sonar, meaning Sound Navigation and Ranging, and which most people have some conceptual familiarity with. LiDAR works by taking a sensor above the terrain and using a pulsed laser to measure the surface. Being that LiDar makes use of a laser, we know the speed of light will be constant and as such can determine the elevation of the reflected pulses as they return to the sensor. The difference in distances means that different elevation models can be created, for example only bare-earth terrain or inclusive of man-made and natural features. This project made use of a basic bare-earth terrain digital elevation model.
The multipatch layer was obtained from the New York City GIS page. A multipatch is an object that is comprised of “patches” that constitute the 3-D shape. Of this data, all of NYC was available but broken up into distinct sections. Since this layer comprised 3-D buildings for every building within the city’s limits, only a specific region was used. This would keep the simplicity of the 3-D model and ease the processing that the applications would have to perform on the data.
The third layer used was a Google Maps screengrab of lower Manhattan. This layer was saved as a .jpg for use in the program. Once loaded into ArcMap, it had to be georeferenced. To georeference an image is to give it a geographic meaning within the GIS software, so it knows where to place the image relative to other geographic locations around the world. Points on the image were selected and given decimal-degree coordinates based on those locations within Google Maps. Only five locations were used and were sufficient to geolocate the image. Although this image was not of the highest resolution considering the scale used, it provided enough of a real-life look to know this is lower Manhattan and give the model more meaning.
Since the area in focus is lower Manhattan and the Google Maps image represents this area, the elevation digital elevation model had to be cut down so that there was not an extraneous amount of data being shown that’s irrelevant. To do this I extracted from the elevation layer exactly the area by using the Google Maps image to cut it out. Finally, a simple layer to depict water was created by outlining the Google Maps image. This layer would be used to model the sea level rise over the elevation, image, and multipatch layers.
Now that I had all of the layers needed, I had to ensure they were all in the same projection. All data used for geospatial analysis must be in the same projection for accurate results. In this case, there was some work to do on the image layer since that had come off Google Maps. The two layers provided by New York City were already in the same Projected Coordinate System (PCS), based on the North American Datum of 1983 (NAD-83). With the image layer, I had to first give it a label so that the program knows the native projection before being able to set the projection to the same as the city layers. Being that Google Maps uses a decimal-degree Geographic Coordinate System (GCS) based on the World Geodetic Datum of 1984 (WGS-84), that is what I set it to. Next, to set it to the same projection as the city layers, I had to find their PCS and used the software’s suggested datum transformation to account for the difference in datums. Once complete, all layers were in the same PCS and ready to be used.
As an aside, the datum is derived from both the geoid, which is an approximation of the Earth’s irregular physical surface and the ellipsoid, which is a smooth rendering of the geoid. By overlaying the ellipsoid on top of the geoid a close approximation of accuracy can be achieved through decimal-degree coordinates based on angles from the center of the ellipsoid. Since there are different ways to position the ellipsoid on top of the geoid, there are different datums that can be made. Some datums are best used for worldwide plotting, such as the WGS-84, and some are best used for certain regions of the world, such as the NAD-83. In the case of the latter, that region would be more accurate in the focused region while sacrificing accuracy in irrelevant regions elsewhere.
Results
At this point, all data was migrated over to ArcScene to build the actual 3-D model. Two views were created. The first view overlayed the elevation and image on top of each other, where I set the symbology of the elevation to a colored scheme that is easily recognizable. The image was set to a transparency of 25% to depict both the Google Maps visualization as well as the colored elevation beneath. To give added effect to the 3-D model, a vertical exaggeration was applied. Since Manhattan is hilly but relatively flat (the highest elevation used in the model is about 170 feet), the exaggeration was set to a factor of 10. The water layer was turned on, set to a transparency of 50%, and colored a deep blue to still be visible. A 10-second animation was created where it blended 20 feet of one-foot intervals of the water layer to show it rising, and then the animation was reversed to show it receding. 20 feet was chosen since this would be best for dramatic effect versus seven feet, not to mention that a 20-foot sea level rise is realistic within the next 200 years.
The second model used the same layers and data but eschewed the vertical exaggeration and added the multipatch layer. The multipatch layer gives another level of conceptual understanding to the real-life effects of sea level rise. Being able to see the buildings with the water rise around them adds that extra bit of realism to help the viewer of the 3-D model grasp the magnitude of the situation.
Conclusion
In all, the 3-D models show us what lower Manhattan will look like when there is a 20-foot rise in sea level. Even if rates of sea level rise are not worst-case-scenario, it is unlikely that we can reverse the fact that sea levels are rising. The most we can do is slow down this rate and hope that future generations can mitigate the risks. Statistics on sea level rise usually cite end-of-century goalposts but humanity will flourish beyond that for, hopefully, millennia to come. A 20-foot rise in sea level is not a matter of “if” but “when.” As such we have an imperative to do more to address the daunting issues that bear down on us, the sooner the better. The least of which is a rising sea level that threatens millions.
