Computers can predict rainfall accurately, scientists say

Computer programmes are far better at accurately predicting rainfall than the government-based seasonal forecasting programmes, Australian scientists say. 

The best monthly rainfall forecasts come from computer programmes able to represent complex relationships between climate data while acquiring knowledge from many examples over time for better pattern detection, said researchers. 

Dr John Abbot and Jennifer Marohasy from the Central Queensland University considered relationships between lagged values for temperature, atmospheric pressure and rainfall as well as climate data. 

They compared results from their artificial neural networks (ANNs) analysis of the Inter-decadal Pacific Oscillation index against government-based seasonal rainfall-forecasting programmes. 

"Forecasts using the ANN for sites in three geographically distinct regions within Queensland are shown to be superior compared to forecasts from the Predictive Ocean Atmosphere Model for Australia (POAMA), which is the general circulation model used to produce the official season rainfall forecasts," researchers said. 

They said a major limitation of government forecasts is that they provide no information about the magnitude of the expected deviation from the median rainfall value within the defined forecast period. 

Researchers said that for purposes including management of water infrastructure or scheduling mine operations, the distribution of rainfall within the three-month period is more important than an averaged seasonal value. 

They said ANNs have been investigated for rainfall forecasting in many parts of the world. 

The study was published in the Atmospheric Research journal published by Elsevier.

Death prediction by Blood test

LONDON: Researchers from Finland and Estonia have identified four biomarkers that can help identify people at high risk of dying from any disease within five years. This was done after screening blood samples from over 17,000 healthy people for over 100 different biomolecules. 

The health status of these study volunteers was followed for several years. The researchers looked for measures in the blood that could reflect who had died within the following 5 years after blood samples were taken. 

"What is especially interesting is these biomarkers reflect the risk for dying from very different types of diseases such as heart disease or cancer. They seem to be signs of a general frailty in the body. Next we aim to study whether some kind of connecting factor between these biomarkers can be identified,'' said researcher Johannes Kettunen. 

In a study published in PLOS Medicine, researchers describe identification of four such biomarkers of death. Of these, albumin was the only one previously linked with mortality. All these molecules are normally present in everyone's blood. 

The novel biomarkers helped detect individuals at much higher risk of dying during five-year follow-up. The measures were independent of well-known risk factors such as age, smoking, drinking, obesity, blood pressure and cholesterol. The result did not change even when only apparently healthy people were examined. 

"We believe that in the future these measures can be used to identify people who appear healthy but in fact have serious underlying illnesses and guide them to proper treatment. More studies are, however, needed before these findings can be implemented in clinical practice," Kettunen said.