Computer-aided design (CAD) significantly enhances the efficiency, accuracy, and innovation of design processes by enabling precise 2D and 3D modeling, extensive analysis, and optimization. Existing methods for creating CAD models rely on latent vectors or point clouds, which are difficult to obtain, and storage costs are substantial. Recent advances in Multimodal Large Language Models (MLLMs) have inspired researchers to use natural language instructions and images for CAD model construction. However, these models still struggle with inferring accurate 3D spatial location and orientation, leading to inaccuracies in determining the spatial 3D starting points and extrusion directions for constructing geometries. This work introduces CAD-GPT, a CAD synthesis method with spatial reasoning-enhanced MLLM that takes either a single image or a textual description as input. To achieve precise spatial inference, our approach introduces a 3D Modeling Spatial Mechanism. This method maps 3D spatial positions and 3D sketch plane rotation angles into a 1D linguistic feature space using a specialized spatial unfolding mechanism, while discretizing 2D sketch coordinates into an appropriate planar space to enable precise determination of spatial starting position, sketch orientation, and 2D sketch coordinate translations. Extensive experiments demonstrate that CAD-GPT consistently outperforms existing state-of-the-art methods in CAD model synthesis, both quantitatively and qualitatively.
Groundwater is an important component in water cycling across agricultural landscapes, but its role in regulating hydrology and nitrogen transport within agricultural watersheds remains highly uncertain, due to limited observations and the reliance on modeling approaches with different model structural representations. Here, we investigated groundwater’s role at the watershed scale in the Upper Sangamon River Basin over 2001–2020 by evaluating its annual and seasonal contributions to both hydrology and nitrogen transport using three validated models with different representations of groundwater-related processes, including (1) default SWAT+, (2) SWAT+gwflow, and (3) SWAT+MODFLOW+RT3D. These models represent increasing levels of groundwater-process complexity within the same model family, enabling us to isolate the effects of groundwater representation on hydrology and nitrogen transport while minimizing differences arising from other model components. We find that groundwater’s annual average net contribution to streamflow ranges from 24.0 % to 31.0 % across models, but its partitioning among contributing pathways varies across models, being dominated by groundwater discharge in the default SWAT+, and by tile flow in SWAT+gwflow and SWAT+MODFLOW+RT3D. The multi-year mean net groundwater contribution to nitrate-N export ranges from 16.0 % to 35.0 %. Seasonal analysis shows that groundwater is a year-round source of nitrate–N to the river system, and its contribution peaks after fertilization at more than 10 times that of the non-growing season. Model comparison shows that physically based groundwater models (SWAT+gwflow and SWAT+MODFLOW+RT3D) better capture surface-subsurface exchange processes, while the empirical groundwater model of default SWAT+ provides efficient flux estimates. Our results highlight the importance of groundwater representation in shaping watershed hydrologic and nitrogen dynamics and underscore the need to balance process realism with computational efficiency when selecting models for agricultural watershed management.
Groundwater is an important component in water cycling across agricultural landscapes, but its role in regulating hydrology and nitrogen transport within agricultural watersheds remains highly uncertain, due to limited observations and the reliance on modeling approaches with different model structural representations. Here, we investigated groundwater’s role at the watershed scale in the Upper Sangamon River Basin over 2001–2020 by evaluating its annual and seasonal contributions to both hydrology and nitrogen transport using three validated models with different representations of groundwater-related processes, including (1) default SWAT+, (2) SWAT+gwflow, and (3) SWAT+MODFLOW+RT3D. These models represent increasing levels of groundwater-process complexity within the same model family, enabling us to isolate the effects of groundwater representation on hydrology and nitrogen transport while minimizing differences arising from other model components. We find that groundwater’s annual average net contribution to streamflow ranges from 24.0 % to 31.0 % across models, but its partitioning among contributing pathways varies across models, being dominated by groundwater discharge in the default SWAT+, and by tile flow in SWAT+gwflow and SWAT+MODFLOW+RT3D. The multi-year mean net groundwater contribution to nitrate-N export ranges from 16.0 % to 35.0 %. Seasonal analysis shows that groundwater is a year-round source of nitrate–N to the river system, and its contribution peaks after fertilization at more than 10 times that of the non-growing season. Model comparison shows that physically based groundwater models (SWAT+gwflow and SWAT+MODFLOW+RT3D) better capture surface-subsurface exchange processes, while the empirical groundwater model of default SWAT+ provides efficient flux estimates. Our results highlight the importance of groundwater representation in shaping watershed hydrologic and nitrogen dynamics and underscore the need to balance process realism with computational efficiency when selecting models for agricultural watershed management.