An Online Approach for Capacity Building: Training Climate Scientists to Use Computer Models

With the mounting evidence by the work of IPCC (2007), climate change has been acknowledged as a significant challenge to Sustainable Development by the international community. In particular, climate extremes, hazards and consequent disasters are imposing additional burden on the world's poorest countries and are also threatening to reverse hard-won development gains. An extreme event has the potential to significantly undermine development, as was the case with Samoa, where two cyclones struck in 1990 and 1991, and set back the country's development by 20 years and wiped out 15 per cent of its Gross Domestic Product (WMO, 2006). Therefore, it is important that scientists in developing countries have access to knowledge and tools (such as climate models) so that well-informed decisions can be made about the mitigation and adaptation of climate change.

However, training these researchers to use new and innovative techniques for climate modeling and data analysis has become a challenge, because current capacity building approaches train researchers to use climate models through short-term workshops (approximately one week), which requires a large amount of funding. In addition, although this type of training is well intentioned, there is a high environmental cost from the greenhouse gas emissions related to traveling to such training workshops. Hence, one could say that capacity building, as it is done today, is expensive, it is not environmentally friendly and it may not be effective in the long run. It has been observed that many participants who recently completed capacity building courses still view climate and weather models as a metaphorical “black box”, where data goes in and results comes out. Additionally, there is evidence that these participants still lack a basic understanding of the climate system. Both of these issues limit the ability of some scientists to go beyond running a model based on rote memorization of the process. As a result, they are unable to solve problems regarding run-time errors, thus becoming dependent on expert modelers. In addition to that, they are not able to assess model limitations in order to make further checks to determine whether or not their model simulation is reasonable (Warner, 2011). Since they view their model as a metaphorical “black box”, it is impossible for them to explore alternatives scenarios in order to improve the simulation.

Current research in the field of science education indicates that there are effective strategies to teach learners about science models. They involve having the learner work with, experiment with, modify, and apply models in a way that is significant and informative to the learner. In the case of computational models, the installation and set up process alone can be time consuming and confusing for new users, which can hinder their ability to concentrate on using, experimenting with, and applying the model to real-world scenarios. It takes time and effort for students to learn about, use, and understand complex models.

Therefore, developing an online version of capacity building is an alternative approach to the workshop training programs. It makes use of new technologies and it allows for a long-term educational process. These long-term training courses also allow participants to engage with the subject matter, either from a theoretical point-of-view or from a practical perspective. It also obviates the need for travel, which makes it very cost-effective and environmentally friendly. In addition to that, online courses can make use of forum discussions, where students can post questions on any doubts they might have, which can lead to discussions between participants as well as instructors. This aspect is extremely important because it has been shown that dialogue that occurs between participants and instructors is a very important component for learning about a new science concept.

A number of science-education courses have already been conducted online within a capacity building project called “The Future of Climate Extremes in the Caribbean (XCUBE)” funded by the Norwegian Directorate for Civil Protection in an assignment for the Norwegian Ministry of Foreign Affairs. If accepted, this presentation will explore a case study related to the online training courses provided via the website m2lab.org for the XCUBE project: “Regional Climate Modeling using WRF”. The course relates to teaching participants how to run WRF for climate simulations using a special version of the model called e-WRF (WRF for Educational purposes) developed by the author. This version of WRF does not require installation and participants can run it on their own laptops, which alleviates the issue of installing and setting up WRF, so that student learning can be focused on using the model itself.

In order to explore the effectiveness of the course, data will be collected from the participants as they complete it. To evaluate content knowledge and student attitudes towards the material and the class, data will be collected from exams, discussion boards, and reflective learning assignments. There are currently 190 participants registered for the course and are made up of graduate students, professors, and researchers from many different science fields. Data is analyzed both quantitatively and qualitatively to assess student learning outcomes from the online course.

Preliminary results indicate that many students enrolled in this course have previously taken a WRF tutorial, but do not feel confident enough to use it, so they have signed up for this course in hopes of gaining the confidence to do regional climate modeling. Despite having taken a tutorial previously, for some participants the basic design of the model was a new concept to them.

For many participants who previously took WRF tutorials, installing WRF so they could access it in their home countries was also problematic. Therefore, participants have expressed their appreciation for the e-WRF, which can be run on a simple desktop. This greatly simplifies the instillation process so that the participants can focus on learning about and using the model.

If accepted, further analysis will be performed as students continue to finish the course. If enough data is available, a statistical analysis will be performed as well as interviews for qualitative data with willing participants.