To investigate the dynamics of stomata, transpiration, and photosynthesis under varying light intensities and CO₂ conditions during leaf development, the light response and CO₂ response of stomatal conductance (g_sw), transpiration rate (T_r), and net photosynthetic rate (P_n) were observed for rice leaves at different days after leaf emergence (DAE). The results showed that (1) as photosynthetically active radiation (PAR) increased, leaf g_sw, T_r, and P_n initially increased rapidly and linearly, followed by a more gradual rise to maximum values, and then either stabilized or showed a declining trend. The maximum g_sw, T_r, and P_n were smaller and occurred earlier for old leaves than for young leaves. The g_sw, T_r, and P_n all exhibited a linear decreasing trend with increasing DAE, and the rate of decrease slowed down with the reduction in PAR; (2) as the CO₂ concentration (C_a) increased, g_sw and T_r decreased gradually to a stable minimum value, while P_n increased linearly and slowly up to the maximum and then kept stable or decreased. The g_sw, T_r, and P_n values initially kept high and then decreased with the increase of DAE. These results contribute to understanding the dynamics in g_sw, T_r, and P_n during rice leaf growth and their response to varied light and CO₂ concentration conditions and provide mechanistic support to estimate dynamic evapotranspiration and net ecosystem productivity at field-scale and a larger scale in paddy field ecosystems through the upscaling of leaf-level stomatal conductance, transpiration, and photosynthesis.

Effective estimation of crop yields at a regional scale holds significant importance in facilitating decision-making within the agricultural sector, thereby ensuring grain security. However, traditional ground-based measurement techniques suffer from inefficiencies, and there exists a need for a reliable, precise, and effective method for estimating regional rice yields. In this study, we employed four machine-learning techniques: partial least squares regression (PLSR), support vector regression (SVR), random forest regression (RFR), and back propagation neural network (BPNN). We combined these methods with multi-temporal rice NDVI (normalized difference vegetation index) data for rice yield estimation. Following an accuracy evaluation and a spatial analysis, the key findings of our study are as follows. (1) The RFR model emerged as the most accurate for rice yield estimation, achieving an R² of 0.65, an RMSE of 388.79 kg/ha, and an rRMSE of 4.48%. While PLSR and SVR demonstrated comparable accuracy, they were both inferior to RFR. (2) Using the top seven predictors with the highest importance rankings as inputs for the RFR model (NDVI values on the 6th, 17th, 33rd, 44th, 71st, 90th, and 106th days after the rice transplanting stage) achieved comparable accuracy while reducing information redundancy. (3) The proposed model demonstrated good spatial applicability (MI = −0.03) for rice yield estimation in Jiangsu, China. (4) A high spatial resolution yearly rice yield dataset (1 km) spanning from 2001 to 2020 was generated using the proposed model and is accessible on the Zenodo database. In conclusion, this study has demonstrated the efficacy of combining multi-temporal remote sensing data with machine-learning techniques for accurate rice yield estimation, thereby aiding agricultural authorities and production enterprises in the timely formulation and refinement of cropping strategies and management policies for the ongoing season.

The photosynthetic light–response (P_n/I) curve is a crucial tool for accurately estimating photosynthetic parameters. However, selecting the most suitable model from numerous available light–response models is still difficult because of the complex canopy structure. This study aimed to evaluate the performance of different models, including the rectangular hyperbolic model (RHM), non-rectangular hyperbolic model (NRHM), exponential model (EM), and modified rectangular hyperbolic model (MRHM), in modeling P_n/I curves and estimating photosynthetic parameters for rice leaves at various canopy positions. The results showed that the NRHM demonstrated the highest accuracy, and the EM was identified as the most ideal in estimating P_n. The RHM consistently overestimated the maximum net photosynthetic rate (P_nmax), apparent quantum efficiency (α), dark respiration rate (R_d), and light compensation point (LCP) while underestimating light saturation point (LSP) for all rice leaves. The NRHM overestimated P_nmax, underestimated LSP, and accurately estimated LCP for all leaves, and overestimated α and R_d for top leaves but performed well for lower leaves. The EM excelled in estimating P_nmax and LSP for all leaves and performed well in estimating α for the top third and fourth leaves, R_d for the top four leaves, and LCP for the top six leaves. The MRHM was effective in estimating P_nmax but consistently overestimated α, R_d, and LSP for all leaves.

Drip irrigation with saline water is frequently adopted to realize the sustainable utilization of saline–sodic soil with high water tables, and soil enzyme activities can be used to indicate changes in soil quality. In the current study, spatiotemporal changes in soil urease enzyme (URE), alkaline phosphatase (ALP) and invertase (INV) activities were investigated during consecutive growing seasons. Soil in beds was sampled before planting (0 y) and one, two, three and four years after the growing season (1 y, 2 y, 3 y, 4 y), and these samples were distributed at four horizontal distances from the drip line (0, 10, 20 and 30 cm) and four vertical soil depths (0–10, 10–20, 20–30 and 30–40 cm). The results showed that a distribution pattern of URE and ALP activities formed during the first growing season, while the distribution of INV activity formed until the third growing season. All three soil enzyme activities in the upper soil layers and positions close to the drip line were more greatly affected by planting year. The average URE activity of the soil profile decreased slightly during the first year and increased by about 220% and decreased by 20% after reclamation for two and three years, and finally, it increased to 4.9 μg NH₄⁺·g⁻¹·h⁻¹ at the end of the fourth growing season. ALP activity remained stable during the first two years and rapidly increased in the following years; in particular, in the fourth year, it reached 32.7 μg ph(OH)·g⁻¹·h⁻¹. INV activity increased continually with the number of years after planting and reached 1009.0 μg glu·g⁻¹·h⁻¹ at the fourth season’s end. An analysis of variance indicated that URE, ALP and INV activities varied insignificantly among the time points of 0 y, 1 y, 2 y and 3 y (p < 0.05), while they were significantly higher for 4 y than for 0 y and 1 y. In addition, all three enzyme activities of the soil profile had an exponentially increasing trend with the number of years after planting. These results indicated the soil quality in saline–sodic soils could be improved with time under drip irrigation with local saline groundwater, especially around the drip line.

Evapotranspiration (ET) plays a crucial role in the surface water cycle and energy balance, and accurate ET estimation is essential for study in various domains, including agricultural irrigation, drought monitoring, and water resource management. Remote sensing (RS) technology presents an efficient approach for estimating ET at regional scales; however, existing RS retrieval algorithms for ET are intricate and necessitate a multitude of parameters. The land surface temperature–vegetation index (LST-VI) space method and statistical regression by machine learning (ML) offer the benefits of simplicity and straightforward implementation. This study endeavors to identify the optimal long-term sequence LST-VI space method and ML for ET estimation under conditions of limited observed variables, (LST, VI, and near-surface air temperature). A comparative analysis of their performance is undertaken using ground-based flux observations and MOD16 ET data. The findings can be summarized as follows: (1) Long-term remote sensing data can furnish a more comprehensive background field for the LST-VI space, achieving superior fitting accuracy for wet and dry edges, thereby enabling precise ET estimation with the following metrics: correlation coefficient (r) = 0.68, root mean square error (RMSE) = 0.76 mm/d, mean absolute error (MAE) = 0.49 mm/d, and mean bias error (MBE) = −0.14 mm. (2) ML generally produces more accurate ET estimates, with the Random Forest Regressor (RFR) demonstrating the highest accuracy: r = 0.79, RMSE = 0.61 mm/d, MAE = 0.42 mm/d, and MBE = −0.02 mm. (3) Both ET estimates derived from the LST-VI space and ML exhibit spatial distribution characteristics comparable to those of MOD16 ET data, further attesting to the efficacy of these two algorithms. Nevertheless, when compared to MOD16 data, both approaches exhibit varying degrees of underestimation. The results of this study can contribute to water resource management and offer a fresh perspective on remote sensing estimation methods for ET.

The distinct co-occurrence of soil water content (SWC) and vapor pressure deficit (VPD) influences ecosystem water use efficiency (WUE) by modifying the synergistic relationship between gross primary productivity (GPP) and evapotranspiration (ET), yet how they impact each other remains unclear in agricultural ecosystems. Based on long-term eddy covariance flux data (2005–2014) observed at a rainfed maize site in Northeast China, we examined how SWC and VPD affect WUE by decomposing it into gross primary productivity to transpiration ratio (GPP/T) and transpiration to evapotranspiration ratio (T/ET). Results showed that WUE was more sensitive to VPD than SWC. Increasing VPD directly suppressed WUE under all soil moisture conditions, whereas SWC had a context-dependent effect: higher SWC reduced WUE under low VPD but enhanced WUE under high VPD. The underlying mechanism was that changes in GPP/T (plant physiological regulation) dominated the WUE responses to both SWC and VPD (contributing 70.25–83.30% and 67.89–87.96%, respectively), while T/ET (evapotranspiration partitioning) played a minor role (<18%). Therefore, to improve WUE under future drier climates, agronomic practices should focus on enhancing photosynthetic capacity and stomatal regulation (e.g., selecting drought-tolerant varieties, optimizing nitrogen supply) rather than solely reducing soil evaporation. Furthermore, supplemental irrigation applied specifically during periods of high VPD (when atmospheric demand is strong) can effectively enhance WUE, as soil moisture becomes critically beneficial under such conditions. These findings provide a mechanistic basis for improving water use efficiency in rainfed maize systems under climate change.