image title

Mapping out the next steps after natural disasters on coastlines

September 13, 2017

ASU geophysicist creates maps from satellite data for first responders during emergencies and policy makers for long-term planning

Knowing the lay of the land is crucial for first responders during emergencies and for civic planners making decisions that direct a city's future. Natural disasters, however, have a way of drastically and suddenly changing that land.

Manoochehr Shirzaei, a geophysicist and radar remote sensing expert at Arizona State University's School of Earth and Space Exploration, specializes in satellites equipped with instruments that use highly accurate remote sensing, called Synthetic Aperture Radar. The data collected from these satellites allow Shirzaei to create high-resolution images of the Earth, with an emphasis on identifying sinking lands and flood zones. The maps are then provided to first responders as well as to public policy makers for long-term planning.

Question: What is a geophysicist, and how does that field relate to flooding caused by hurricanes like Harvey and Irma?

Answer: Geophysics is the study of the physical processes and physical property of Earth (and its surrounding space environment) by mapping the variations of physical property that are remotely sensed. In my case, I’m using satellites to study the Earth’s surface and specifically ground-level changes, which can be the result of natural causes or human-caused ones like extraction for water or fuel. 

Q. You created a map of the Houston area with flooding caused by Hurricane Harvey (the map is pictured above, with flooded areas in red). How was this map created?

A: We acquired data from the Sentinel 1A/B satellites that belong to the European Space Agency. The instruments on these satellites use radar to provide highly accurate remote sensing data. There are two satellites that have a six-day revisit time (orbiting the Earth every six days), so we can compare an area before and after a disaster. 

The raw images at first look like black and white dots, as if you were to spread large amounts of salt and pepper over a sheet of paper.  Once we’ve processed the data, we can colorize the mapped areas based on levels of flooding. 

See a larger version of the map here.

Q: How will the map be used?

A: During an intense storm, it is difficult to fly an airplane or drone over an affected area. And then clouds often are in the way of any satellite pictures that would show us what the flooding may look like. But radar can get through both clouds and rain.

This is helpful for first responders so they can determine where aid and relief is needed the most. It can also help with estimating the overall damage of an area. 

In terms of forecasting, remote sensing is also useful to determine which land is above or below sea level and therefore more prone to flooding. 

A map showing areas flooded in Florida after Hurricane Irma

A map of southeast Florida flooding caused by Hurricane Irma generated from two C-Band SAR images acquired by Sentinel 1A/B satellites between Aug. 28 and Sept. 10. Image by Manoochehr Shirzaei

Q: And how did you create a similar map for the flooding caused by Hurricane Irma?

A: I contacted the European Space Agency in advance of the storm hitting Florida and requested data from their Sentinel 1A/B satellites. Once the storm hit and the data were available for the affected area, I began creating maps depicting the areas of flooding along the Florida coast. These maps are then provided to the appropriate local and national authorities so they can better assist those areas in need. 

Q: You recently were recently selected to join the NASA Sea Level Change Team. What will you be doing for NASA as part of this team?

A: I will be working on mapping the U.S. coasts and studying the coastal land subsidence as well as the impact of sea-level rise on coastal flooding. NASA will use this data to inform local and national public authorities so that they can plan for flooding and infrastructure improvements, with the goal to minimize future damage.  

Top photo: Map of Houston flooding caused by Hurricane Harvey generated from two C-Band SAR images acquired by Sentinel 1A/B satellites between Aug. 24 and 30. Image by Manoochehr Shirzaei

Karin Valentine

Media Relations & Marketing manager , School of Earth and Space Exploration


image title

How can apps get users to generate content? ASU study finds gender differences

Competition or cooperation as motivator? Men and women differ, ASU study finds.
September 13, 2017

Research team discovers that cooperation, competition are different motivators

China’s largest recipe-sharing platform needed a carrot to motivate more content from users, and research from a team of Arizona State University professors was able to pinpoint what works.

A new study by faculty in the W. P. Carey School of Business found that specific kinds of notifications could elicit more content from the app users — and that there are differences between men and women. Feedback that promoted a message of helping others prompted women to contribute more content, while men were more likely to respond to competitive messages.

The findings are an example of use-inspired research — important because many huge companies are dependent on user-created content, and they are constantly looking for ways to prompt customers to contribute.

Ni Huang, an assistant professor of information systems at ASU and the lead author, said that the teamThe other authors are Bin Gu, a professor of information systems, and Chen Liang, a doctoral student, both from the W. P. Carey School of Business at ASU, and Gordon Burtch, an assistant professor in the the Carlson School of Management at the University of Minnesota. Gu is the Earl and Gladys Davis Distinguished Professor and is associate dean for China programs at the W. P. Carey School of Business. worked with Meishi, a Chinese company that owns the largest recipe-sharing smartphone application in China, in which users contribute and rate one another’s recipes.

Also on the research team was Yili Hong, an assistant professor of information systems at ASU who has done previous research on companies that require user-generated content.

“One of the questions that resonated with me was that they have tens of millions users but guess how many contribute? Less than 10 percent,” he said of the recipe app.

The Chinese company agreed to let the researchers perform the study with all the users in the new “foodie talk” section of the platform, and the 1,129 subjects were divided into four groups. Each group received a different kind of push alert on their smartphones once a week for seven weeks, and then their content contributions for the following week were recorded.

App users received different types of push notifications.

The control group received a generic message inviting users to look at the app, Huang said.

“Most of the time we find that kind of message is not very effective compared to when you provide performance feedback, how well you are doing in terms of the content you’ve contributed,” she said.

The other groups received feedback on how many “likes” they had, but the content was framed in different ways. One group’s push alerts gave the number of “likes” and emphasized how much they helped other people, saying, “You have provided cooking inspiration” for X number of other users.

Another group received a push alert giving a ranking, such as “you are in the top 3 percent of users.” This is “individualist” framing.

The fourth group was competitive, comparing performance, such as “you beat 98 percent of other foodies.”

The results differed by gender. In the group that received the “helpful” messages, everyone produced more content, but the effect was bigger for women, who contributed about 6 percent more postings than the control group. The “helpful” feedback prompted males to contribute about 2 percent more postings than the control.

In the competitive group, males generated nearly 11 percent more content than the males in the control group — but females responded negatively, uploading about 2 percent fewer postings than females in the control group.

“For females, if you tell them they’re outperforming other people, they’ll actually contribute less,” Hong said.

There was little difference between males and females who received the “individualist” messages with a ranking, Huang said.

The study also compared engaged users, who were already contributing a lot of good content, with users who contributed but did not receive a lot of “likes.”

“Think of it like ‘good students’ and ‘poor students,’ ” Hong said. “When we give this performance feedback, the good students will be more responsive and do even better, but when we tell the poor students this feedback, they’re less responsive. It’s demotivating.”

Companies that rely on user content already know that feedback generates more content among users, and they try a variety of methods to incentivize them.

“But no one knows which practice works or how to optimize for different users and genders, and that is the main contribution of this paper,” Hong said.

When the team submitted the paper to the journal Management Science, the editors were concerned that because the experiment involved cooking, an activity that could be considered female-oriented, the results might not be generalizable. So the group did another crowd-sourcing experiment, using 1,000 online subjects in the United States to rate (non-cooking-related) content, and replicated the results. The paper was accepted and published Aug. 31. Find the study here.

The Chinese app company was excited about the results and asked the team to do more research.

“If you think about the results, we were only looking at people who are already contributing some sort of content, and getting some likes,” Hong said. “But if a person has never contributed anything, how can you convert them into someone who is engaged? Is there a way to nudge them?”

The researchers are looking at the concept of fairness: “We’re using that concept to say, ‘You have benefitted from others; why don’t you try something yourself?’ ” They hope to produce another paper on that topic.

Mary Beth Faller

Reporter , ASU Now