Accurate precipitation forecasting is crucial for mitigating the impacts of ex-treme weather events and enhancing disaster preparedness.This study evalu-ates the performance of Long Short-Term Memory and Bidirectional...Accurate precipitation forecasting is crucial for mitigating the impacts of ex-treme weather events and enhancing disaster preparedness.This study evalu-ates the performance of Long Short-Term Memory and Bidirectional LSTM models in predicting hourly precipitation in Dar es Salaam using a multivariate time-series approach.The dataset consists of temperature,pressure,U-wind,V-wind,and precipitation,preprocessed to handle missing values and normal-ized to improve model performance.Performance metrics indicate that BiLSTM outperforms LSTM,achieving lower Mean Absolute Error and Root Mean Squared Error by 6.4%and 6.5%,respectively along with improved threshold scores.It demonstrated better overall prediction accuracy.It also im-proves moderate precipitation detection(TS3.0)by 16.9%compared to LSTM.These results highlight the advantage of bidirectional processing in capturing complex atmospheric patterns,making BiLSTM a more effective approach for precipitation forecasting.The findings contribute to the development of im-proved deep learning models for early warning systems and climate risk man-agement.展开更多
Air pollution is among the most serious environmental and public health problems worldwide,especially in low and middle-income countries like Rwanda.This study explores the spatial and temporal variations of criteria ...Air pollution is among the most serious environmental and public health problems worldwide,especially in low and middle-income countries like Rwanda.This study explores the spatial and temporal variations of criteria air pollutants across Rwanda from 2019 to 2023,utilizing data from 18 national air quality monitoring stations and 16 weather stations.Results reveal that PM2.5 and PM10 concentrations exceeded WHO guidelines,with the mean reaching 90μg/m3(PM2.5)and 127μg/m3(PM10),predominantly in Kigali City,Northern,and Western provinces.CO concentration peaked in the Eastern province and Kigali.In contrast,NO_(2) and O3 were highest in the Central and Northern provinces.Over five years,NO_(2) showed a slight increase trend,while CO,O3,and SO_(2) displayed minor declines and remained in line with WHO guidelines.Diurnal variations highlighted morning(06:00-07:00 am)and evening(06:00-09:00 pm)pollutant peaks,driven by morning rush hour traffic,domestic stoves,and industrial activities.Border stations like Bugeshi-Rubavu recorded elevated pollutant levels due to cross-border emissions from the bordering countries.Seasonal analysis revealed higher pollutant levels during dry seasons,influenced by reduced rainfall and increased anthropogenic activities.CO concentration was positively correlated with temperature during MAM(r=0.69)due to increased biomass burning and agricultural emissions.Wind speed is negatively correlated with PM2.5 and PM10 in JJA,aiding pollutant dispersion,while PM2.5 is positively correlated with humidity in MAM(r=0.7),linked to secondary aerosol formation.These findings underscore the urgent need to improve air quality,particularly in urban and border regions,and address Rwanda’s transboundary pollution concerns.展开更多
文摘Accurate precipitation forecasting is crucial for mitigating the impacts of ex-treme weather events and enhancing disaster preparedness.This study evalu-ates the performance of Long Short-Term Memory and Bidirectional LSTM models in predicting hourly precipitation in Dar es Salaam using a multivariate time-series approach.The dataset consists of temperature,pressure,U-wind,V-wind,and precipitation,preprocessed to handle missing values and normal-ized to improve model performance.Performance metrics indicate that BiLSTM outperforms LSTM,achieving lower Mean Absolute Error and Root Mean Squared Error by 6.4%and 6.5%,respectively along with improved threshold scores.It demonstrated better overall prediction accuracy.It also im-proves moderate precipitation detection(TS3.0)by 16.9%compared to LSTM.These results highlight the advantage of bidirectional processing in capturing complex atmospheric patterns,making BiLSTM a more effective approach for precipitation forecasting.The findings contribute to the development of im-proved deep learning models for early warning systems and climate risk man-agement.
文摘Air pollution is among the most serious environmental and public health problems worldwide,especially in low and middle-income countries like Rwanda.This study explores the spatial and temporal variations of criteria air pollutants across Rwanda from 2019 to 2023,utilizing data from 18 national air quality monitoring stations and 16 weather stations.Results reveal that PM2.5 and PM10 concentrations exceeded WHO guidelines,with the mean reaching 90μg/m3(PM2.5)and 127μg/m3(PM10),predominantly in Kigali City,Northern,and Western provinces.CO concentration peaked in the Eastern province and Kigali.In contrast,NO_(2) and O3 were highest in the Central and Northern provinces.Over five years,NO_(2) showed a slight increase trend,while CO,O3,and SO_(2) displayed minor declines and remained in line with WHO guidelines.Diurnal variations highlighted morning(06:00-07:00 am)and evening(06:00-09:00 pm)pollutant peaks,driven by morning rush hour traffic,domestic stoves,and industrial activities.Border stations like Bugeshi-Rubavu recorded elevated pollutant levels due to cross-border emissions from the bordering countries.Seasonal analysis revealed higher pollutant levels during dry seasons,influenced by reduced rainfall and increased anthropogenic activities.CO concentration was positively correlated with temperature during MAM(r=0.69)due to increased biomass burning and agricultural emissions.Wind speed is negatively correlated with PM2.5 and PM10 in JJA,aiding pollutant dispersion,while PM2.5 is positively correlated with humidity in MAM(r=0.7),linked to secondary aerosol formation.These findings underscore the urgent need to improve air quality,particularly in urban and border regions,and address Rwanda’s transboundary pollution concerns.
