The highly complex memristive chaotic map provides an excellent alternative for engineering applications. To design a memristive chaotic map with high complexity, this paper proposes a new three-dimensional memristive chaotic map (named MLM) by cascading and coupling a non-locally active memristor with a locally active memristor. The dynamical behaviors of MLM are revealed through phase diagrams, Lyapunov exponent spectra, bifurcation diagrams, and dynamic distribution diagrams. Notably, the internal frequency of MLM exhibits unique LE-controlled behavior and shows an extension of the chaotic parameter range. The high complexity of MLM is validated through the use of Spectral entropy (SE) and C0, and Permutation Entropy (PE) complexity algorithms. Subsequently, a pseudorandom number generator (PRNG) based on MLM is designed. NIST test results validate the high randomness of the PRNG. Finally, the STM32 hardware platform is used to implement MLM, and attractors under different parameters are measured by an oscilloscope, verifying the numerical analysis results.

Accurate monthly temperature and precipitation forecasting in Guizhou Province is challenging because of its complex karst terrain and substantial spatiotemporal variability in meteorological conditions. We propose a forecasting framework that integrates Seasonal-Trend Decomposition using Loess (STL), Variational Mode Decomposition (VMD), bidirectional Long Short-Term Memory (BiLSTM) encoding, dynamic graph learning, graph convolution, and spatiotemporal cross-attention. STL and VMD provide multiscale temporal features, which are combined with meteorological, temporal, and station-related variables for spatiotemporal representation learning. An auxiliary high-precipitation loss is introduced to improve the prediction of high-precipitation events. Experiments were conducted using monthly observations from 38 meteorological stations in Guizhou Province during 1961-2011 under a chronological, leakage-free evaluation protocol. The proposed model achieved an RMSE of 1.759 °C for temperature and 57.550 mm for precipitation. Although it did not obtain the lowest aggregate precipitation error among the evaluated baselines, it achieved an upper-tail precipitation RMSE of 93.096 mm and MAE of 73.053 mm lower than those of monthly climatology, STGCN, and Graph WaveNet. Moving-block bootstrap analysis further supported the reductions in upper-tail RMSE and MAE relative to Graph WaveNet. These results indicate that the primary advantage of the proposed framework lies in forecasting high-precipitation magnitudes rather than minimizing aggregate precipitation error alone.

Weather forecasting in complex terrain, such as the karst landscapes of Guizhou, faces the dual challenge of capturing microscale topographic effects and macroscale atmospheric dynamics. Existing models often struggle to balance local spatiotemporal patterns with long-range dependencies. To address this, we present a novel hybrid architecture that integrates ConvLSTM and Transformer via a Spatiotemporal Cross-Attention (STCA) mechanism. Our approach utilizes an STL-VMD preprocessing pipeline to decompose non-stationary meteorological signals into stationary subsequences, effectively separating trend and seasonal components. The architecture synergizes ConvLSTM's local feature extraction with Transformer's global correlation modeling, while the STCA module dynamically bridges spatial and temporal information through multi-scale convolution and adaptive gating. Validated across 33 meteorological stations in Guizhou over a 51-year period (1961-2011), our model demonstrates superior performance. Compared to baseline LSTM models, temperature forecasting achieves an 18.99% reduction in RMSE ([Formula: see text]), and notably, precipitation forecasting achieves a 27.74% reduction in RMSE with a substantial 54.08% improvement in [Formula: see text]. The results confirm the model's robustness and generalization capability for operational forecasting in topographically complex regions.