The use of automated feature extraction for diagnosing double inter-tropical convergence zones

We develop an automated and user-trained algorithm that spatially analyzes satellite data and detects the presence of double inter-tropical convergence zones (DITCZ) and apply it to 30 years worth of satellite collected outgoing long-wave radiation data. The analysis of the data via vertical wavelet transform identifies DITCZ occurrences over the Indian Ocean in the last three decades.

► We develop an automated, user-trained algorithm to spatially analyze satellite data.
► The algorithm detects the presence of double inter-tropical convergence zones (DITCZ).
► We apply algorithm to 30 years of satellite outgoing long wave radiation data.
► Analysis via wavelet transform identifies previously undiagnosed Indian Ocean DITCZs.