Tajikistan’s Future Climate: Using Statistics to “See” the Future

The Pamir Mountains, one of the main geographic features of Tajikistan. Photo: dwrawlinson/Flickr

By Michael McCormick, Climate and Society ’13

The threat of a changing climate is imminent, requiring actions coming from various industries. In the climate science industry, people rely on forecasting to predict what sorts of events can occur in the immediate and long-term future of the planet. These forecasts are integral pieces to citizens across the globe who can use the data to change certain habits and minimize negative impacts from a changing world.

In order to create these forecasts, several pieces of information are needed, including accurate historical temperature and precipitation records. But perhaps more important for analyzing these records is a strong background in statistics. Using various statistical methods, such as linear regressions and correlation analyses, we as scientists are able to communicate to a variety of people in relatively simple terms what are the possible outcomes of this ever-changing climate.

During my time in the Climate and Society program, there was a strong emphasis on understanding the climate system – from small scale to global circulation patterns. But one class in particular directly enabled my fellow colleagues and me to translate scientific knowledge to actual climate outcomes, all through the use of statistics. Our climate modeling class combined common statistical tools with what we all were interested in, climate change. We learned how to use statistical analyses to breakdown large datasets of precipitation, temperature, and crop yield.

During my summer work with Dr. Brad Lyon of the International Research Institute for Climate and Society (IRI), I have utilized the statistical skills gained in the climate modeling class, including using the expansive IRI Data Library to locate large climate datasets. I have also begun using the Climate Predictability Tool (CPT) developed by the IRI that aids in the statistical analysis of these large datasets.

My area of interest is in the seasonal climate predictability for the country of Tajikistan. The country has a complex history, including a civil war and fall of the former Soviet Union, destroying most of their meteorological stations. With such sparse data in the region, there is a need for accurate data and predictions, as Tajikistan has been cited as one of the most climate-vulnerable countries in the world. The country and its people could use help in understanding the climatic changes that are likely to occur, and seasonal predictability will help facilitate behavioral changes in natural resource management.

The lack of historical data is concerning for climate researchers in the area, as many predictions and forecasts cannot rely on the inaccurate historical record. However, using multiple regression and correlation analyses acquired in the Climate and Society climate modeling course, I can more accurately predict seasonal climate. And, for such a vulnerable country to climate change, the predictions and possible climate outcomes are incredibly important to Tajiks.

To me, there is somewhat of a duty for climate scientists to not only conduct groundbreaking research, but to create and distribute tangible, useful results. If we as climate scientists are unable to help those who don’t have the resources we do, we fail. Using the worldwide language of statistics and math, we can convey possible outcomes due to climate change to those who cannot find out themselves. Across the globe, including Tajikistan, people can – and will – use these seasonal predictions to alter natural resource management techniques, and it is our duty as climate scientists to provide them these tools.