Abstract Leaf Area Index (LAI) is a crucial biophysical parameter in terrestrial ecosystem and hydrological modeling. Although LAI intercomparisons have been extensively conducted at coarse- to medium-resolution scales, systematic multi-site evaluations of high-resolution (≤30 m) LAI products remain limited. This research compared three high-resolution LAI products from National Ecological Observatory Network (NEON) Airborne Observatory Platform Imaging Spectrometer (NIS), Sentinel-2 (S2), Global Ecosystem Dynamics Investigation (GEDI). These LAI products were also evaluated by Digital Hemispheric Photos (DHPs) measurements. Our findings show a greater consistency between GEDI and S2 ( R 2 = 0.31 at the pixel level, and R 2 = 0.69 at the site level) than that of GEDI and NIS ( R 2 = 0.12 at the pixel level, and R 2 = 0.51 at the site level). The comparison of S2 and NIS reaches a similar level of agreement at both pixel level ( R 2 = 0.26) and site level ( R 2 = 0.78). Time series analysis of GEDI and S2 LAI finds a reasonably good agreement on tracking canopy phenology. Comparisons with DHPs total LAI suggest that both S2 and GEDI can achieve a good estimation at site level. Our results suggest the potential of fusing GEDI and S2 data for improved high-resolution forest LAI mapping and monitoring despite their differences in instrument and sampling design. • Intercomparison of high-resolution LAI from multispectral, hyperspectral, and lidar sensors. • NEON airborne hyperspectral LAI shows the overall least agreement with reference datasets. • Sentinel-2 LAI and GEDI LAI have a good agreement at site level. • GEDI LAI shows the overall best agreement with in-situ total LAI.

Accurate forecasting of water demand is essential for reliable and efficient operation of distribution networks. However, existing forecasting approaches are usually restricted to single-site closed settings and fail under distributional shifts across pumping stations, leaving the gap of cross-site generalization unresolved. This study addresses the problem using operational data collected from Supervisory Control and Data Acquisition (SCADA) systems provided by the Public Utility Company for Water Treatment Valjevo, which operates the Kolubara water supply region in Serbia. We extend the Informer architecture with statistical alignment (CORAL, MMD), adversarial adaptation (DANN), and meta-learning (MAML, Reptile) to explicitly handle zero , few , and full shot transfer scenarios under covariate shift. The proposed framework reduces mean and peak errors across forecasting horizons, improving resilience where operations are most vulnerable to demand surges, and thus carries direct social relevance for water security. Results show that statistical/adversarial alignment enables effective zero -shot transfer, while meta-learning supports rapid adaptation with only 24–72 h of labeled data, consistently outperforming classical and deep-learning baselines. The evaluation, conducted on multivariate SCADA data, was checked using standard accuracy and peak-sensitive metrics. Generally, the study establishes cross-site transfer learning as an engineering solution for water utilities, offering a deployable pipeline that adapts to new pumping stations with minimal calibration while reducing operational risks and costs and grounding its value in both methodological innovation and empirical validation. • Cross-site forecasting framed as domain adaptation under covariate shift. • Informer backbone extended with CORAL, MMD, DANN, MAML, and Reptile variants. • Meta-adaptation enables rapid few -shot learning with only 24 to 72 labeled hours. • Residual ACF/PACF diagnostics confirm removal of domain-specific autocorrelation. • Supports water security through adaptive forecasting across heterogeneous pumping sites.