Better weather and climate forecasting

We have never had so much information about the Earth’s atmosphere and the lands and oceans as now, but there are still large gaps in this knowledge. These gaps, combined with the imperfections of forecasting models and the complexity of climate processes, prevent accurate understanding of the current and future state of the atmosphere and the climate system.

Weather forecasts depend on the accuracy of the initial conditions and the numerical models of the atmosphere. The analyzes are carried out on the basis of measurements of temperature, wind speed and humidity made at a height of about 60 – 80 km above the Earth’s surface. Currently, the biggest disadvantage of the global observation system is the lack of accurate information about the winds, especially over the oceans.

Numerical models of NWP, based on mathematical simulations of processes in the atmosphere, allow to predict weather taking into account current data from observations of weather conditions. These models allow forecasting the weather for several hours or even for a whole week ahead in various spatial scales. For the correct climate prediction, however, these models must contain both algorithms for forecasting conditions in the atmosphere, over the oceans and on land, as well as formulas that reflect the composition of the atmosphere and its fluctuations. Modeling results are widely used to learn about current and future climate change.

The lack of observational data and the imperfections of the models make analyzes and forecasts have a certain degree of uncertainty. However, even with the most perfect models and well-known initial conditions, the longer the forecasted period, the greater the uncertainty of the forecast – all due to the natural non-linearity of flows.