Hybrid deep learning framework for evaluating field evapotranspiration considering the impact of soil salinity

Entitled “A novel hybrid deep learning framework for evaluating field evapotranspiration considering the impact of soil salinity” for possible publication in Water Resources Research.

Data:

The Salinized Farmland Flux Station sites are located in the typical irrigated agricultural area of the arid continental monsoon region in northwest China.In this study, we used data from four flux tower located in saline farmland: two for maize (MZ1, MZ2) and two for sunflower (SF1, SF2).

For each site, we collected the following variables at half-hourly temporal resolution: (i) latent heat (LE, W m^-2) fluxes, serving as a direct measure representing the energy component of ET (mm h^-1), (ii) net radiation (R_n, W m^-2), (iii) ground heat flux (G, W m^-2), (iv) solar irradiance (R_s, W m^-2), (v) air temperature (T_a, °C), (vi) vapor pressure deficit (VPD, KP_a), (vii) wind speed (U_s, m s^-1), (viii) relative humidity (RH, %), and (ix) atmospheric carbon dioxide concentration (C_a, mg m^-3). During the crop growth period, field in-situ measurements of soil and vegetation data from these farmlands are conducted approximately every 10 days.  

Code:

All the codes were executed in Python. The provided hybrid deep learning model code can be run on Jupyter Notebook.