This research takes the Suzhou Gardens, a World Cultural Heritage Site, as the object of study and investigates a rapid scheme generation approach for garden restoration and expansion projects, assisting designers in making scientific decisions. Considering the limitations of current garden design, which is inefficient and relies on human experience, this study proposes an intelligent generation framework based on a conditional generative adversarial network (CGAN). In constructing the CGAN model, we determine the spatial characteristics of the Suzhou Gardens and, combined with historical floor plan data, train the network. We then design an optimization strategy for the model training process and finally test and verify the generative space scheme. The research results indicate the following: (1) The CGAN model can effectively capture the key elements of the garden space and generate a planar scheme that conforms to the traditional space with an accuracy rate reaching 91.08%. (2) This model can be applied to projects ranging from 200 to 1000 square meters. The generated results can provide multiple scheme comparisons for update decisions, helping managers to efficiently select the optimal solution. (3) Decision-makers can conduct space utilization analyses and evaluations based on the generated results. In conclusion, this study can help decision-makers to efficiently generate and evaluate the feasibility of different design schemes, providing intelligent support for decision-making in urban renewal plans.

This study focuses on the surface materials of rammed-earth walls of Fujian Tulou in Xiaoshu Village, exploring the microscopic characteristics of rammed earth in different orientations and the microclimate adaptation mechanism and degradation law of the walls. Specimens were collected from the inner and outer surface soil layers of the four directional walls of a representative Tulou. SEM, XRD, and XRF analyses were performed to characterize the materials’ microstructure, mineral composition, and elemental distribution, with the test results correlated to the microclimatic conditions of each wall orientation. The conclusion is as follows: (1) The microscopic particle size of rammed earth exhibits significant directional differences at dual scales of 300 nm and 2 μm. Solar radiation duration and wind speed are positively correlated with the coefficient of variation in particle size. (2) The southeast and north walls were the most severely damaged (soil loss, quartz enrichment: 79.9%), the west wall had minor cracks, the north wall showed slight salt crystallization (Halite = 0.3%), and the east wall exhibited moisture-related moss growth. (3) Traditional organic additives (bamboo strips, rice husks) mitigate deterioration and enhance structural integrity. (4) The diversity of soil color (related to hematite and iron oxide) can serve as a simple indicator of deterioration. This study has proposed differentiated protection schemes for the “microclimate-compounds” on walls facing different directions on the rammed-earth surface of the Tulou. The findings provide a theoretical basis for orientation-specific conservation of Tulou heritage and offer scientific references for the modification of modern rammed-earth materials.

At present, the raindrop size distribution (DSD) fitting method and traditional neural network method are the main quantitative precipitation estimation (QPE) algorithms. The former is not enough to express the spatiotemporal variability of the DSDs using generalized $Z-R$ parameter relationship, and the latter is limited by the number of network layers, network structure and computing power. Their performance cannot both meet actual needs. In this research, we propose an alternative approach to dual-polarization radar (DPR) QPE. In particular, a radar-gauges dataset (RGD) is constructed for QPE based deep learning using radar raw data and gauges measurements for landfall typhoon in South China, and a deep learning model is designed and trained using this dataset. The model is applied to radar data to produce rainfall estimates. Preliminary results show the promising performance of this novel method compared to traditional QPE estimators.

The long-term preservation of earthen wall heritage in Shanxi largely depends on the stability of its rammed earth walls. In this study, the ancient city wall site of Guoyang in Shanxi is taken as a case study. Samples are collected from four directional sections (brick, mortar, and rammed earth), and a nonlinear dynamic finite element method (FEM) with horizontal impact loading is applied, combined with microscopic analyses (SEM, XRF, XRD), to investigate the deterioration mechanisms and structural responses of the brick wall surface. The results show that (1) under identical stress, the north wall exhibits smaller displacement (2.942 × 10⁻² mm) than the south wall (4.058 × 10⁻² mm), indicating greater lateral stiffness and deformation resistance; (2) the rammed earth particles on the north side show the highest uniformity (CV = 0.4981), while those on the west side display the greatest dispersion (CV = 0.9852), revealing pronounced differences in microscopic stability among the different wall orientations; (3) the north-side rammed earth exhibits a composition typical of high-quality traditional ternary lime–soil mixtures, characterized by high SiO₂ (47.06%), high CO₂ (19.73%), and low Na₂O (2.358%); and (4) the durability of the four wall sections varies with distinct deterioration characteristics. Mineral composition and environmental factors jointly contribute to the differences in wall color and surface conditions. This study provides a reference for the conservation and restoration of similar rammed earth heritage sites.