Bridging the gap between GRACE and GRACE-FO missions with deep learning aided water storage simulations

Uz, Metehan; Atman, Kazım Gökhan; Akyılmaz, O.; Shum, C. K.; Keleş, Merve; Ay, Tuğçe; Tandoğdu, Bihter; Zhang, Yu; Mercan, H.

doi:10.1016/j.scitotenv.2022.154701

Cited by 24 publications

(18 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is well known that surface water and groundwater have completely different dynamic behaviours. The change in the surface water is much faster than the change in groundwater (Uz et al, 2022).…”

Section: Discussionmentioning

confidence: 99%

A fuzzy logic framework to handle uncertainty in remote sensing‐based hydrological data for water budget improvement across mid‐ and small‐scale basins

Kayan

Türker

Erten

2022

Hydrological Processes

View full text Add to dashboard Cite

In recent years, the efficacy of remote sensing (RS) products for water budget (WB) analysis has been widely tested and implemented in global and regional basins. Although RS products provide high temporal and spatial resolution images with near‐global coverage, uncertainty is still a significant problem. Fuzzy logic is a powerful technique for dealing with uncertainty in different engineering problems. In this study, the annual residual error (r$$ r $$) in the WB equation arising from the uncertainties of the RS products was minimized by applying fuzzy correction coefficients to each WB component. For analysis, three different fuzzy linear regression (FLR) models with 14 different sub‐models were used in the two basins having different spatial characteristics, namely Sakarya and Cyprus basins. Although FLR sub‐models produce similar findings in the Sakarya basin, they generated more complex results in the Cyprus basin. This is mainly due to the higher uncertainty of the RS products in the Cyprus basin. The Cyprus basin is too small for some low‐resolution RS‐based products to resolve, and it has a higher leakage error due to across ocean/land boundary. In addition, the general performance of sub‐models is better in the Sakarya basin than that in the Cyprus basin. The best fuzzy sub‐models reduced the error up to 68% and 52% in terms of mean absolute error compared with the non‐fuzzy model in the Sakarya and Cyprus basins, respectively. Further evaluations showed that the best sub‐model precipitation well captured the temporal patterns of gauge observations in both basins. Moreover, they have the best consistency with gauge observations in terms of root mean square error, Kling‐Gupta efficiency, and percent bias in both basins. The results proved that this study will provide valuable insights into WB analysis in ungauged basins by incorporating the fuzzy logic approach into hydrological RS products.

show abstract

Section: Discussionmentioning

confidence: 99%

A fuzzy logic framework to handle uncertainty in remote sensing‐based hydrological data for water budget improvement across mid‐ and small‐scale basins

Kayan

Türker

Erten

2022

Hydrological Processes

View full text Add to dashboard Cite

show abstract

“…Compared with the GRACE/GRACE‐FO solution, the Swarm measurements have excessive noise in higher harmonics (Teixeira da Encarnação et al., 2020). To suppress this noise, we truncated this Swarm gravity field at d/o 12 (Uz et al., 2022) and applied a 750 km Gaussian filter recommended by Teixeira da Encarnação et al. (2020) to the truncated field.…”

Section: Methodsmentioning

confidence: 99%

A Two‐Step Linear Model to Fill the Data Gap Between GRACE and GRACE‐FO Terrestrial Water Storage Anomalies

Yang,

You,

Tian

et al. 2023

Water Resources Research

View full text Add to dashboard Cite

The Gravity Recovery and Climate Experiment (GRACE) and its Follow‐On (GRACE‐FO) missions have revolutionized global terrestrial water storage anomalies (TWSA) measurements. However, the 11‐month data gap between the two GRACE missions disrupts the measurement continuity and limits its further applications. Previous attempts to fill this data gap require further improvement in terms of method robustness and product quality. Here, we propose a novel two‐step linear model using precipitation, temperature data, and hydrological model‐simulated TWSA as predictors to fill the 11‐month data gap between the two GRACE missions and generate six global gridded GRACE‐like TWSA products from April 2002 to July 2021. These products are evaluated at grid scale globally and also basin scale for the world's largest 72 river basins. Results indicate that our GRACE‐like data show great consistency with the GRACE/GRACE‐FO observations. While most basins exhibit consistent performance across the six GRACE‐like TWSA products, certain areas with lower signal‐to‐noise ratios show significant variability. Furthermore, we assess the performance of our GRACE‐like data during the data gap using one previous reconstruction, a hydrological model simulation, and the Swarm satellite measurement. The results confirm that our GRACE‐like data exhibit equivalent performance within and outside the data gap. This study introduces a more simple and robust method for predicting the missing data between the two GRACE missions and provides readily applicable continuous GRACE‐like TWSA products for hydrologic applications.

show abstract

“…The GRACE mission operated from 2002-2017, followed by a gap of 1 year, after which GRACE-FO was launched. Several studies have filled the data gap using deep learning (e.g., Uz et al, 2022), and availability of modeled GRACE data from such studies can be used to extend the study for a longer time. This is beyond the scope of this study and left for a future work.…”

Section: Limitations and Future Studiesmentioning

confidence: 99%

Machine learning based downscaling of GRACE-estimated groundwater in Central Valley, California

Agarwal

Akyılmaz

Shum

et al. 2023

Science of The Total Environment

Self Cite

View full text Add to dashboard Cite

Bridging the gap between GRACE and GRACE-FO missions with deep learning aided water storage simulations

Cited by 24 publications

References 50 publications

A fuzzy logic framework to handle uncertainty in remote sensing‐based hydrological data for water budget improvement across mid‐ and small‐scale basins

A fuzzy logic framework to handle uncertainty in remote sensing‐based hydrological data for water budget improvement across mid‐ and small‐scale basins

A Two‐Step Linear Model to Fill the Data Gap Between GRACE and GRACE‐FO Terrestrial Water Storage Anomalies

Machine learning based downscaling of GRACE-estimated groundwater in Central Valley, California

Contact Info

Product

Resources

About