Surface and sub-surface flow estimation at high temporal resolution using deep neural networks

Recent intensification in climate change have resulted in the rise of hydrological extreme events. Hydrological processes modelling at high temporal resolution is required to better understand flow patterns at catchment scale. A physically-based model called Hydrological Simulated Program-FORTRAN and two deep learning-based models were implemented to model surface runoff and sub-surface flow in the tropical humid headwater catchment of Houay Pano, northern Lao PDR. One deep learning model consisted of only one long short-term memory (simple LSTM), whereas the other model simulated processes in each hydrological response unit (HRU) by defining one separate LSTM for each HRU (HRU-based LSTM). The models use time-series data and two-dimensional spatial data to predict surface and sub-surface flows at 6-minute time step. The simple LSTM model outperformed the other models on surface runoff prediction, whereas HRU-based LSTM model better predicted sub-surface flow. This study demonstrates the performance of a deep learning model when simulating hydrological cycle with high temporal resolution.

The paper is an achievement of the GETiEES ParisUNIST collaboration and was published in the Journal of Hydrology. 50 days’ free access is available here.

More news

Reuse of bottom sediment from reservoirs to cropland is a promising agroecological practice that must be rationalized

In semi-arid areas, intermittent streams are often equipped with small reservoirs to store water for irrigation and/or groundwater recharge, and to capture sediments lost through erosion. These reservoirs must be periodically desilted to maintain their storage capacity. While bottom sediments are generally considered waste, their reuse in agricultural fields is a centuries-old practice in India. […]

Soil erosion control in tree plantations on steep slopes: Runoff water andsediment trapping efficiency of riparian grass buffer in mountainoushumid tropics

Riparian grass buffers reduce the velocity of water flowing over the soil surface during storms, capturing surface runoff (SR) and trapping soil particles eroded from cultivated slopes. Rarely quantified under steep slope conditions (>45 %), this phenomenon probably occurs in many mountain agroecosystems in the humid tropics. In Southeast Asia, teak plantations are often established on […]

Village settlements in mountainous tropical areas, hotspots of fecal contamination as evidenced by Escherichia coli and stanol concentrations in storm water pulses

Little is known about the contribution of villages to the fecal contamination of surface water in tropical rural areas. This study, published in the journal Environmental Science & Technology, documents E. coli and stanol concentration levels in surface runoff, and tracks the origin of the fecal contamination during flood events.

Search