GIS 4048, Storm tracking and coastal flooding
This was a great lab, in part because I feel like I'm synthesizing a lot of skills I've learned previously into GIS skills, and not just following a lab guide.
The first part was making a storm tracker for Hurricane Sandy.
The storm tracker section was fairly straightforward, but the trickiest parts were using the world countries layers and US boundaries layers to have unique labeling for the Sandy affected areas. This involved me labeling the US boundaries but excluding the FEMA states with a SQL expression, and making a new layer of Sandy affected countries from the world countries layer and labeling that, then labeling world countries with a SQL expression that excluded labeling the new layer.
I also went a step beyond the lab guide and added a label for the first record of each day so I could show a temporal aspect. I got really stuck here, because the provided DATE___TIM column only shows the date in the attribute table, but it was labeling MM/DD/YYYY and adding HH:MM:SS time format on the end. Apparently this is some kind of default on ArcGIS, and I wound up using AI to help me make an Arcade expression to label it and exclude the time of day.
At some point, I accidentally labeled the storm track line, and had an incongruous giant block label, and spent the better part of an hour turning on and off layers, on and off labels, and hitting undo/redo before figuring out what it was. I also got deep in the weeds of playing with font size, labeling strategies, color, halo, basemap, etc.
Part 2 of the lab was creating a citizen reporter survey for damage assessment (not included here). I’ve used things like SurveyMonkey before, but I had a hard time figuring out the interface per the lab instructions. I eventually just watched ESRI tutorials on YouTube.
Finally, the storm damage assessment was fun. We created domains for storm damage (a rating of how much structural damage, wind damage, inundation, etc.) and used that to create points to label parcels of a beach area affected by Hurricand Sandy. Creating domains is useful, if a little tedious. I wonder if these could be created in Excel and imported, or some other way to avoid clicking through dozens of fields and manually entering. I used multiple ring buffer and intersect to make layers with each damage category filtered by distance, which is one of those moments where I actually feel like I’m starting to really get and internalize GIS skills.
The readings this week talked about the amount of agreement between volunteer assessors on damage assessment like this, and how it varies between high and low volume respondents, and professionals and volunteers. There are probably hundreds of current and former students who have done this lab, it would be interesting to look at how much agreement we have. Comparing myself to other blogs in the class, it looks like I was very much on the low side of assessed damage.
The first part was making a storm tracker for Hurricane Sandy.
The storm tracker section was fairly straightforward, but the trickiest parts were using the world countries layers and US boundaries layers to have unique labeling for the Sandy affected areas. This involved me labeling the US boundaries but excluding the FEMA states with a SQL expression, and making a new layer of Sandy affected countries from the world countries layer and labeling that, then labeling world countries with a SQL expression that excluded labeling the new layer.
I also went a step beyond the lab guide and added a label for the first record of each day so I could show a temporal aspect. I got really stuck here, because the provided DATE___TIM column only shows the date in the attribute table, but it was labeling MM/DD/YYYY and adding HH:MM:SS time format on the end. Apparently this is some kind of default on ArcGIS, and I wound up using AI to help me make an Arcade expression to label it and exclude the time of day.
At some point, I accidentally labeled the storm track line, and had an incongruous giant block label, and spent the better part of an hour turning on and off layers, on and off labels, and hitting undo/redo before figuring out what it was. I also got deep in the weeds of playing with font size, labeling strategies, color, halo, basemap, etc.
Part 2 of the lab was creating a citizen reporter survey for damage assessment (not included here). I’ve used things like SurveyMonkey before, but I had a hard time figuring out the interface per the lab instructions. I eventually just watched ESRI tutorials on YouTube.
Finally, the storm damage assessment was fun. We created domains for storm damage (a rating of how much structural damage, wind damage, inundation, etc.) and used that to create points to label parcels of a beach area affected by Hurricand Sandy. Creating domains is useful, if a little tedious. I wonder if these could be created in Excel and imported, or some other way to avoid clicking through dozens of fields and manually entering. I used multiple ring buffer and intersect to make layers with each damage category filtered by distance, which is one of those moments where I actually feel like I’m starting to really get and internalize GIS skills.
The readings this week talked about the amount of agreement between volunteer assessors on damage assessment like this, and how it varies between high and low volume respondents, and professionals and volunteers. There are probably hundreds of current and former students who have done this lab, it would be interesting to look at how much agreement we have. Comparing myself to other blogs in the class, it looks like I was very much on the low side of assessed damage.


Comments
Post a Comment