COIF-NN: Cloud Onset Instant Foreteller - Neural Network

Predicting cloud/fog onsets with ground-based (FTIR-) radiometer data via COIF-NN
COIF-NN - cloud onsets' instant foreteller - neural network

 
EXAMPLES OF PREDICTION

Lidar dataPredicted gap evolution, SGP, June 30 2002
Lidar data & Predicted gap evolution
"Gap"= difference between actual- and dewpoint-temperature profiles: an onset occurs when  the  "gap" reaches "0"
Prediction has started at  9:00 GMT, driven by the operator of evolution which has been estimated  with rather noisy  data (see, e.g. parameters of observed spectra below, yet examples given with the special page ) over the time window   7:30 - 9:00 GMT.
Note, that the lidar backscatter  data  indicate a cloud onset  at  1.5 km about 12:00 hours GMT., which has not  been developed into a  cloud  with a well-detected  by the lidar cloudbase, but effected the thermodynamic evolution  of the atmospheric boundary layer beneath.

Prediction of  IR- spectrum evolution  (the fisrt 3 spectral parameters are displayed)
Gennady Ryzhikov
                                                                         aka Геннадий Рыжиков
To the previous COIF-page. to the previous COIF-page
to the next COIF-page to the next COIF-page