Breaks in monsoon and related precursors

A comprehensive analysis of eleven break monsoon situations that occurred during the period 1987 to 1997 have been attempted in the study. The various features like daily rainfall departures, wind anomalies and the satellite derived Outgoing Long wave Radiation (OLR) associated with the commencement/cessation of the break monsoon condition are studied with a view to identifying the precursors associate the break situation. The results reveal that there is progressive decrease of below normal rainfall departures 5 days prior to the actual break day in the latitude belts south of 20° N. During the period of the revival of the monsoon, the time section of the daily rainfall departures shows that the daily rainfall departure first starts becoming above normal in the southern most latitudinal belt 5° N to 10°N from the second day onwards after the cessation of the break. Similarly, the easterly anomalies in the zonal wind are first noticed in the southern latitude even 5 days prior to the starting of the break in the lower and middle troposphere. The maximum easterly anomalies in the lower and the middle troposphere move northwards upto 20° N. The composite latitudinal time section of OLR anomaly show a large area of negative OLR anomaly extending from 20°S to 10°N. The area is defined as the Southern. Hemispheric Convective Zone ( SHCZ). The negative OLR anomaly (10 Wm-2 is noticed around 5° S to 0° N. It increases to 20 Wm-2 on the second day of the break on the same latitudinal belt. The daily OLR anomaly pattern shows that the area of the negative OLR anomaly around the equatorial region increases with the approach of a break epoch. The forecasting aspects of the commencement / cessation of the break have been also discussed.

Anomaly Pattern Detection in Streaming Data Based on the Transformation to Multiple Binary-Valued Data Streams

Anomaly pattern detection in a data stream aims to detect a time point where outliers begin to occur abnormally. Recently, a method for anomaly pattern detection has been proposed based on binary classification for outliers and statistical tests in the data stream of binary labels of normal or an outlier. It showed that an anomaly pattern can be detected accurately even when outlier detection performance is relatively low. However, since the anomaly pattern detection method is based on the binary classification for outliers, most well-known outlier detection methods, with the output of real-valued outlier scores, can not be used directly. In this paper, we propose an anomaly pattern detection method in a data stream using the transformation to multiple binary-valued data streams from real-valued outlier scores. By using three outlier detection methods, Isolation Forest(IF), Autoencoder-based outlier detection, and Local outlier factor(LOF), the proposed anomaly pattern detection method is tested using artificial and real data sets. The experimental results show that anomaly pattern detection using Isolation Forest gives the best performance.

Cold water temperature anomalies on the Sodwana reefs and their driving mechanisms

The Sodwana reef system experiences short-term temperature fluctuations that may provide relief from bleaching and be crucial in the future survival of the system. These temperature fluctuations are best described as cold water temperature anomaly events that occur over a period of days and cause a drop in temperature of a few degrees on the reef. We explored the statistical link between the temperature anomalies and the regional hydrodynamics to elucidate the driving mechanisms of the temperature anomalies around Sodwana. Temperature measurements taken between 1994 and 2015 on Nine‑Mile Reef at Sodwana show that temperature anomalies occur on average three times per year at Sodwana and predominantly during the summer months. A conditional average of altimetry data at the peak of the temperature anomalies showed the emergence of a negative sea surface height (SSH) anomaly pattern and associated cyclonic eddy just offshore of the Sodwana region. The cyclonic eddies associated with the temperature anomalies originate on the southwestern edge of Madagascar and migrate westwards until they interact with the African coastline at Sodwana. Instantaneous altimetry SSH fields over the 21-year period were cross-correlated to the conditionally averaged SSH field within a 2° region around Sodwana. It was found that 33% of the temperature anomalies at Sodwana were not associated with the presence of cyclonic eddy systems. This finding suggests that an offshore cyclonic eddy interacting with the shelf is not the sole driving mechanism of the temperature anomalies.

Decadal Variation of the Precipitation Relationship Between June and August over South China and Its Mechanism

This paper investigates the characteristics and causes for the interdecadal change in the relationships between early and late summer rainfall over South China (SC). This study finds that the correlations of the precipitation over SC between June and August shift from weakly positive in 1979 – 1995 to obviously negative in 1996-2019. Further analysis demonstrates that the interdecadal variations of monthly SST anomaly (SSTA) and associated air-sea interactions in June and August account for the decadal variations of the precipitation relationships. During the prior period 1979-1995, the tropical West Indian Ocean (WIO) shows a significant positive SSTA in June, which triggers Kevin waves and an anticyclone circulation over the tropical Northwest Pacific (NWP). The warm and wet air transported by the southwest airflow at the north of the anticyclone provides favorable environmental condition to produce more precipitation over SC region in June. In contrast, the SST dipole pattern with the negative SSTAs in the maritime continent (MC) and positive SSTAs in the tropical Central Pacific (CP) is dominant in August. The SST dipole pattern is inconducive to the formation of anticyclone over SC, causing a weak positive precipitation correlation between June and August. During the latter period 1996-2019, the precipitation over SC in June is the same as that in the prior period as there is no significant decadal change in tropical WIO SST and East Asian circulation. However, an opposite phase of the SST dipole anomaly pattern in MC and the tropical CP is dominant in August during the latter period. Accordingly, the positive feedback mechanism of air-sea interaction leads to the enhancement of local convection activities in MC and the meridional Hadley circulations and the NWP subtropical high, leading to a decrease of precipitation over SC in August. Overall, the decadal variation of the SST dipole anomaly pattern in MC and the tropical CP is the key factor affecting the adjustment of the correlations between June and August precipitation in the two periods.

Rainfall Anomalies Pattern in Northwestern Nigeria (NWN)

Constant investigation into rainfall anomaly pattern is very crucial as it enables the detection of any departure from normal rainfall condition. When such departure is persistent and statistically significant, it could indicate climate change. This study seeks to investigate anomaly pattern of rainfall in north western Nigeria with the view to determine any extreme departure from established normal rainfall behavior (mean). The study used thirty years (30) rainfall data from 1987 to 2016. The data was obtained from the archives of Nigerian Meteorological Agency (NIMET) for six selected synoptic stations from the region. Purposive sampling technique was adopted in selecting the six synoptic stations given consideration to stations with longer consistent rainfall records. The data was subjected to Standardized Anomalies also known as Standardized Precipitation Index (SPI) to obtain anomaly values. The values were used to plot time series for each station. They were also used to determine the dry or years with drought i.e. negative values and wet or moisture years i.e. positive values. The findings showed that throughout the thirty years period, normal conditions dominated the study area with few pockets of dry conditions. The study concludes that rainfall anomalies pattern in north western Nigeria over the thirty years period under investigation was not too far from normal rainfall conditions.

Characterization of the sulfide deposits in the southeastern Nigeria using VLF method: insights from numerical modeling and field examples

A priori geologic and geophysical information has been used to construct conceptual VLF experiments on conductively and inductively coupled overburden geological models of the lead-zinc (Pb-Zn) mineralization zone found in southeastern Nigeria. This is based on the finite element approach to (1) simulate different geologic situations of overburden occurrence, (2) examine the roles played by overburden in modifying and masking VLF responses of a buried conductor target, and (3) confirm the effectiveness of VLF method in mapping lead-zinc lodes found in sedimentary terrains. The computed theoretical model curves and field examples are expected to serve as guide for VLF anomaly pattern recognition due to overburden thickness, resistivity and width of conductor in similar terrain as the study area.

Graph Deep Learning for Long Range Forecasting

<p>Deep learning-based models have been recently shown to be competitive with, or even outperform, state-of-the-art long range forecasting models, such as for projecting the El Niño-Southern Oscillation (ENSO). However, current deep learning models are based on convolutional neural networks which are difficult to interpret and can fail to model large-scale dependencies, such as teleconnections, that are particularly important for long range projections. Hence, we propose to explicitly model large-scale dependencies with Graph Neural Networks (GNN) to enhance explainability and improve the predictive skill of long lead time forecasts.</p><p>In preliminary experiments focusing on ENSO, our GNN model outperforms previous state-of-the-art machine learning based systems for forecasts up to 6 months ahead. The explicit modeling of information flow via edges makes our model more explainable, and it is indeed shown to learn a sensible graph structure from scratch that correlates with the ENSO anomaly pattern for a given number of lead months.</p><p> </p>