Exploring the Potential of Statistical Modeling to Retrieve the Cloud Base Height from Geostationary Satellites: Applications to the ABI Sensor on Board of the GOES-R Satellite Series.

Although cloud base height is a relevant variable for many applications, including aviation, it is not routinely monitored by current geostationary satellites. This is probably a consequence of the difficulty of providing reliable estimations of the cloud base height from visible and infrared radiances from current imagers. We hypothesize that existing algorithms suffer from the accumulation of errors from upstream retrievals necessary to estimate the cloud base height, and that this hampers higher predictability in the retrievals to be achieved. To test this hypothesis, we trained a statistical model based on the random forest algorithm to retrieve the cloud base height, using as predictors the radiances from Geostationary Operational Environmental Satellites (GOES-16) and variables from a numerical weather prediction model. The predictand data consisted of cloud base height observations recorded at meteorological aerodrome report (METAR) stations over an extended region covering the contiguous USA. Our results indicate the potential of the proposed methodology. In particular, the performance of the cloud base height retrievals appears to be superior to the state-of-the-science algorithms, which suffer from the accumulation of errors from upstream retrievals. We also find a direct relationship between the errors and the mean cloud base height predicted over the region, which allowed us to obtain estimations of both the cloud base height and its error.

Examining the Potential of a Random Forest Derived Cloud Mask from GOES-R Satellites to Improve Solar Irradiance Forecasting.

In order for numerical weather prediction (NWP) models to correctly predict solar irradiance reaching the earth's surface for more accurate solar power forecasting, it is important to initialize the NWP model with accurate cloud information. Knowing where the clouds are located is the first step. Using data from geostationary satellites is an attractive possibility given the low latencies and high spatio-temporal resolution provided nowadays. Here, we explore the potential of utilizing the random forest machine learning method to generate the cloud mask from GOES-16 radiances. We first perform a predictor selection process to determine the optimal predictor set for the random forest predictions of the horizontal cloud fraction and then determine the appropriate threshold to generate the cloud mask prediction. The results show that the random forest method performs as well as the GOES-16 level 2 clear sky mask product with the ability to customize the threshold for under or over predicting cloud cover. Further developments to enhance the cloud mask estimations for improved short-term solar irradiance and power forecasting with the MAD-WRF NWP model are discussed.