Machine learning algorithms for merging satellite-based precipitation products and their application on meteorological drought monitoring over Kenya

Ghosh, Suravi; Lu, Jianquan; Das, P. N. Mohan; Zhang, Zhenke

doi:10.1007/s00382-023-06893-6

Cited by 5 publications

(11 citation statements)

References 78 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This maintains good data quality for drought computation. Previous studies extensively used CRU-TS datasets for evaluating the satellite precipitation products and global climate models for extreme event monitoring 7 , 11 , 25 , 59 , 60 . Thus, this study uses CRU-TS data to evaluate the SRPPs and HEML precipitation estimates for drought monitoring.…”

Section: Experimental Design and Methodologymentioning

confidence: 99%

“…The ERA5-Land reanalysis dataset is produced through an Integrated Forecasting System (IFS) using a 4D-var assimilation algorithm in version 41r2 by the European Centre for Medium-Range Weather Forecast (ECMWF) 67 . The datasets have been available from 1950 to the present with high spatiotemporal resolution (0.1° × 0.1°) at a global scale 11 . The ERA5-L products contain different climatic parameters, including precipitation estimates.…”

Section: Experimental Design and Methodologymentioning

confidence: 99%

“…In the future, regions accustomed to high precipitation levels may unexpectedly experience drought due to climate and land use changes, posing significant challenges for water resource management and environmental protection 7 – 9 . During the preceding decade (2003–2012), drought events adversely impacted 36.5 million individuals, leading to a substantial loss of life across the globe 10 , 11 . Additionally, the drought impacts 30% of the world's population and leads to huge economic losses 12 .…”

Section: Introductionmentioning

confidence: 99%

“…However, it should be noted that the sampling error, different retrieved algorithms, and various sensors are the major cause of biases in SRPPs 41 . Hence, various algorithms have been created to amalgamate these SRPPs, aiming to enhance the precision of precipitation estimates 11 , 37 .…”

Section: Introductionmentioning

confidence: 99%

“…Zandi et al 46 proposed a stacked generalisation ensemble approach consisting of RF, SVM, and MLP (Multilayer perceptron) neural network for developing new high-resolution precipitation estimates and compared to LWLR (local weighted linear regression) approach in a topographically complex region. In a recent investigation conducted by Ghosh et al 11 , dual Machine Learning (ML) algorithms, including Gradient Boosting Machine (GBM), Random Forest (RF), Support Vector Machine (SVM), and k-nearest Neighbors (KNN), were employed for the purpose of monitoring drought conditions in Kenya. However, most of the previous studies only focused on developing new precipitation estimates using different ML models and did not justify the capacity of these precipitation data for extreme climate event detection, especially drought in East Africa.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

A hybrid ensemble learning merging approach for enhancing the super drought computation over Lake Victoria Basin

Das,

Zhang,

Ghosh

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

This study introduces a novel Hybrid Ensemble Machine-Learning (HEML) algorithm to merge long-term satellite-based reanalysis precipitation products (SRPPs), enabling the estimation of super drought events in the Lake Victoria Basin (LVB) during the period of 1984 to 2019. This study considers three widely used Machine learning (ML) models, including RF (Random Forest), GBM (Gradient Boosting Machine), and KNN (k-nearest Neighbors), for the emerging HEML approach. The three SRPPs, including CHIRPS (Climate Hazards Group Infra-Red Precipitation with Station), ERA5-Land, and PERSIANN-CDR (Precipitation Estimation from Remotely Sensed Information using Artificial Neural Network-Climate Data Record), were used to merge for developing new precipitation estimates from HEML model. Additionally, classification and regression models were employed as base learners in developing this algorithm. The newly developed HEML datasets were compared with other ML and SRPP products for super-drought monitoring. The Standardized precipitation evapotranspiration index (SPEI) was used to estimate super drought characteristics, including Drought frequency (DF), Drought Duration (DD), and Drought Intensity (DI) from machine learning and SRPPs products in LVB and compared with RG observation. The results revealed that the HEML algorithm shows excellent performance (CC = 0.93) compared to the single ML merging method and SRPPs against observation. Furthermore, the HEML merging product adeptly captures the spatiotemporal patterns of super drought characteristics during both training (1984–2009) and testing (2010–2019) periods. This research offers crucial insights for near-real-time drought monitoring, water resource management, and informed policy decisions.

show abstract

Section: Experimental Design and Methodologymentioning

confidence: 99%

Section: Experimental Design and Methodologymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A hybrid ensemble learning merging approach for enhancing the super drought computation over Lake Victoria Basin

Das,

Zhang,

Ghosh

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

Evaluating hydrological responses of satellite precipitation products over an Indian tropical catchment through a distributed physical model

Kalura,

Pandey,

Chowdary

et al. 2024

Hydrological Processes

View full text Add to dashboard Cite

Significant advancements in satellite‐based precipitation retrieval algorithms have led to the development of continuous, quasi‐global precipitation products, offering unique opportunities for hydrometeorological and climate research. In this context, six satellite‐based precipitation products (SPPs), including CHIRPS, CMORPH, GSMaP, IMERG, MSWEP, and PERSIANN, were thoroughly investigated for their application in hydrological simulations over the Wardha River basin in India. The observed gridded precipitation product developed by India Meteorological Department (IMD) has been used as reference data to evaluate the performance of SPPs. Hydrological variables such as runoff, soil moisture (SM) and evapotranspiration (ET) were simulated using the Variable Infiltration Capacity model. The performances of SPPs are critically assessed using various statistical metrics, including Pearson's correlation coefficient (R), Klinga‐Gupta Efficiency (KGE), Percent Bias (PBIAS), root mean square error (RMSE), and the RMSE to standard deviation ratio (RSR). The model‐simulated discharge was compared with streamflow observations at a single gauging site, while a spatially distributed comparison was conducted between simulated SM and ET and satellite‐based SM and ET. The IMD dataset consistently shows superior performance for discharge simulation at both daily and monthly scales, with KGE values of 0.85 and 0.91, respectively. Among SPPs, MSWEP and CHIRPS excel in simulating daily and monthly discharge, respectively. For ET simulation, CHIRPS outperforms IMD and other SPPs, achieving an overall KGE value of 0.37. In contrast, PERSIANN is the most effective for simulating SM compared to other precipitation products.

show abstract

An innovative method for measuring the hysteresis effects of soil moisture on meteorological variables at various time scales and climate conditions

Tian,

Lu,

Chen

2023

Geo-spatial Information Science

View full text Add to dashboard Cite

Machine learning algorithms for merging satellite-based precipitation products and their application on meteorological drought monitoring over Kenya

Cited by 5 publications

References 78 publications

A hybrid ensemble learning merging approach for enhancing the super drought computation over Lake Victoria Basin

A hybrid ensemble learning merging approach for enhancing the super drought computation over Lake Victoria Basin

Evaluating hydrological responses of satellite precipitation products over an Indian tropical catchment through a distributed physical model

An innovative method for measuring the hysteresis effects of soil moisture on meteorological variables at various time scales and climate conditions

Contact Info

Product

Resources

About