Evaluation of Remote-Sensing Reflectance Products from Multiple Ocean Color Missions in Highly Turbid Water (Hangzhou Bay)

Validation of remote-sensing reflectance (Rrs) products is necessary for the quantitative application of ocean color satellite data. While validation of Rrs products has been performed in low to moderate turbidity waters, their performance in highly turbid water remains poorly known. Here, we used in situ Rrs data from Hangzhou Bay (HZB), one of the world’s most turbid estuaries, to evaluate agency-distributed Rrs products for multiple ocean color sensors, including the Geostationary Ocean Color Imager (GOCI), Chinese Ocean Color and Temperature Scanner aboard HaiYang-1C (COCTS/HY1C), Ocean and Land Color Instrument aboard Sentinel-3A and Sentinel-3B, respectively (OLCI/S3A and OLCI/S3B), Second-Generation Global Imager aboard Global Change Observation Mission-Climate (SGLI/GCOM-C), and Visible Infrared Imaging Radiometer Suite aboard the Suomi National Polar-orbiting Partnership satellite (VIIRS/SNPP). Results showed that GOCI and SGLI/GCOM-C had almost no effective Rrs products in the HZB. Among the others four sensors (COCTS/HY1C, OLCI/S3A, OLCI/S3B, and VIIRS/SNPP), VIIRS/SNPP obtained the largest correlation coefficient (R) with a value of 0.7, while OLCI/S3A obtained the best mean percentage differences (PD) with a value of −13.30%. The average absolute percentage difference (APD) values of the four remote sensors are close, all around 45%. In situ Rrs data from the AERONET-OC ARIAKE site were also used to evaluate the satellite-derived Rrs products in moderately turbid coastal water for comparison. Compared with the validation results at HZB, the performances of Rrs from GOCI, OLCI/S3A, OLCI/S3B, and VIIRS/SNPP were much better at the ARIAKE site with the smallest R (0.77) and largest APD (35.38%) for GOCI, and the worst PD for these four sensors was only −13.15%, indicating that the satellite-retrieved Rrs exhibited better performance. In contrast, Rrs from COCTS/HY1C and SGLI/GCOM-C at ARIAKE site was still significantly underestimated, and the R values of the two satellites were not greater than 0.7, and the APD values were greater than 50%. Therefore, the performance of satellite Rrs products degrades significantly in highly turbid waters and needs to be improved for further retrieval of ocean color components.

Forecasting Rainfall in Mauritius using Seasonal Autoregressive Integrated Moving Average and Artificial Neural Networks

In this paper, two forecasting methods namely, the autoregressive integrated moving average (ARIMA) and the artificial neural network (ANN) are studied to forecast the amount of rainfall in Mauritius. Indeed due to the geographical location of Mauritius, the rainfall pattern is deeply affected by the season prevailing whereby the period of summer receives a relatively high amount of rainfall when compared to winter. As such, forecasting rainfall can help the local authorities to manage the distribution of water in the country especially during droughts. The results obtained from both methods are compared in terms of their mean square error, mean absolute difference and mean absolute percentage difference. It is then seen that artificial neural network is a much better model as it is more accurate. This is due to its nonlinearity characteristic and ability to learn and train itself.

Accuracy of Ultrasound Estimated Fetal Weight in Small for Gestational Age and Appropriate for Gestational Age Grown Periviable Neonates

Objective Clinicians use estimated fetal weight (EFW) as a proxy for birth weight (BW) in the antenatal period. Our objective was to compare the accuracy of EFW obtained by ultrasound to BW among infants born during the periviable period and determine if accuracy of EFW varied among small for gestational age (SGA) versus appropriate for gestational age (AGA) grown neonates. Study Design We included women who delivered between 230/7 and 256/7 weeks' gestation and had an EFW within 7 days of delivery. Mean percentage difference and median absolute percentage difference between EFW and BW were calculated. Results Our cohort included 226 neonates with a mean gestational age of 241/7 ± 0.8 weeks and median BW of 653 g (interquartile range [IQR]: 580–750 g). The median absolute percentage difference between EFW and BW of fetal weight estimates was 9.2% (IQR: 3.6–17.2). EFW overestimated BW for 75% (n = 171) of the cohort. Among SGA infants, the mean percentage difference in EFW and BW was 16.2 ± 19.4% versus 6.9% ± 13.1% in AGA infants (p = 0.019). Conclusion EFW overestimated BW in this cohort. In addition, ultrasound was less accurate among infants born SGA. These data are important to consider when counseling families facing periviable delivery.