Reliable climate information can help countries plan for adverse and beneficial climate events, allocate resources, adapt to climate change and achieve development goals.
Advances in climate science, including forecasting on seasonal and sub-seasonal timescales, real-time climate monitoring, creation of decadal-scale and climate change information, and tailoring of climate information to specific user needs, are creating opportunities to improve climate risk management and adaptation to climate change.
Our activities include real-time global seasonal forecasts, forecast development, diagnostics and modeling research, and tool development and capacity building. We work closely with sectoral experts in health, water resources, agriculture and disaster management, to identify areas in which climate information can be used for decision making and planning, and to ensure that the information is communicated in the best way and tailored to the needs of the users.
Real-Time seasonal forecasts
Each month IRI issues a General Circulation Model (GCM) multi-model based seasonal forecast of global precipitation and temperature. Partners provide seasonal climate model outputs, which are used as essential input to IRI’s forecasts. The IRI detects and corrects individual GCM systematic biases from forecasts of past years, and then consolidates the GCM forecasts into a single probability forecast. Forecast products include global maps showing regions of increased probability for being wetter-, dryer-, warmer- or colder-than-normal for the upcoming seasons, also available in flexible formats. El Nino Southern Oscillation (ENSO) forecasts are also issued monthly in the form of the IRI/Climate Prediction Center (CPC) ENSO Prediction Plume, and ENSO Quick-Look, with input from over 20 institutions worldwide.
A good forecast is one that is not only scientifically sound but is also usable, fits into a user’s decision system, is accessible (e.g. via IRI Data Library), and is trusted, emphasizing the importance of forecast verification. Our strategic elements are the forecast inputs/ingredients (climate prediction models, observational data); key products, which include categorical probabilities and full probability distribution functions of the forecast distribution; methodologies, including model combination/calibration which occurs between the model predictions and putting out forecast maps; and product delivery, which is largely by means of the IRI Data Library and its various “maprooms.” Our current work includes incorporating more coupled GCMs, including the US National Multi-Model Ensemble (NMME), into the seasonal forecast; new more flexible forecast products that leverage emerging research, and MME methodology that is more agile, automated, and able to adapt to changing GCM model inputs. Our recent work also includes subseasonal forecast development using similar multi-model approaches.
Diagnostics and modeling research
We undertake strategic climate research to address fundamental climate questions that arise in regional settings, build institutional credibility and research partnerships in the US and beyond. Diagnostic analysis can help build climate science capacity in developing countries where we are working, and provide applications of more fundamental research from the climate science community. Examples of our diagnostic studies with practical implications include the seasonally varying influence of ENSO in the Philippines; using climate information to develop indicators and understanding the processes involved in natural decadal variability and regional climate change in places like southeast South America and the Sahel; and analyses of seasonal predictability of local-scale sub-seasonal rainfall quantities like daily rainfall frequency and mean rainfall intensity.
Tool development and capacity development
Our aim is to assist those responsible for providing operational climate information, such as National Meteorological Services and Regional Climate Centers in developing countries. We assist them in their climate service responsibilities through a wide range of capacity-development activities including trainings, tool development, and establishment of best practices and standards. For example, the Climate Predictability Tool (CPT) is designed to assist National Meteorological Services to produce their own tailored, downscaled seasonal climate forecasts, either using global datasets (such as sea temperature measurements) or dynamical model outputs from the World Meteorological Organization’s Global Producing Centres.