Duy Minh Dao

Nghiên cứu nhằm đánh giá hiện trạng phân bố nước mặt theo đơn vị hành chính và làm rõ sự khác biệt về tỷ lệ diện tích nước mặt giữa các xã, phường trên địa bàn tỉnh Thái Nguyên, qua đó cung cấp cơ sở phục vụ công tác quản lý và quy hoạch phát triển. Nghiên cứu sử dụng ảnh Landsat 8 OLI và chỉ số nước khác biệt chuẩn hóa (Normalized Difference Water Index - NDWI) để nhận diện, xác định diện tích nước mặt; sau đó ứng dụng Hệ thống thông tin địa lý để tính tỷ lệ diện tích nước mặt theo từng đơn vị hành chính và phân loại theo phương pháp phân khoảng tự nhiên. Kết quả cho thấy tài nguyên nước mặt phân bố không đồng đều: 85,87% số xã, phường thuộc mức rất thấp, 8,70% thuộc mức thấp và chỉ 5,43% đạt từ mức trung bình đến rất cao. Đại Phúc có tỷ lệ diện tích nước mặt lớn nhất, đạt 13,115%, tiếp đến là Vạn Phú và Đại Từ. Cách tiếp cận này sử dụng dữ liệu mở, chi phí thấp, dễ cập nhật và phù hợp với quản lý tài nguyên nước, quy hoạch phát triển ở tỉnh Thái Nguyên.

Mapping of tea plantations is vital for agricultural planning in Thai Nguyen city, Thai Nguyen province, Vietnam. This study applies an improved classification method using Landsat 8 imagery, integrating spectral reflectance, Normalized Difference Vegetation Index (NDVI), and phenological characteristics on the Google Earth Engine platform. Imagery from April and December was utilized to capture seasonal NDVI variations that enhanced land cover classification. In December, tea plantations exhibit lower NDVI values (0.38–0.47) due to dormancy, clearly separating them from evergreen forests (0.49–0.52) and recently harvested rice fields. In contrast, April imagery shows overlapping NDVI ranges among tea, crops, and forests, complicating classification. Using Landsat 8 December images, combining NDVI, spectral reflectance and seasonal ecological characteristics (phenology) in a machine learning algorithm analysis framework significantly improves the efficiency of land cover classification. The classification was performed using Support Vector Machine (SVM)and Random Forest (RF) algorithms, with the RF model achieving 91.3% overall accuracy and a Kappa coefficient of 0.89, outperforming SVM by 2.2%. These findings confirm the effectiveness of combining spectral features and NDVI with the Random Forest algorithm for accurate tea plantation mapping.

Tea is one of the most significant cash crops and plays an important role in economic development and poverty reduction. On the other hand, tea is an optimal choice in the extreme weather conditions of Tanuyen Laichau, Vietnam. In our study, the NDVI variation of tea in the growing season from 2009 to 2018 was showed by calculating NDVI trend and the Mann-Kendall analysis to assess trends in the time series. Support Vector Machine (SVM) and Random Forest (RF) model were used for predicting tea yield. The NDVI of tea showed an increasing trend with a slope from −0.001–0.001 (88.9% of the total area), a slope from 0.001–0.002 (11.1% of the total area) and a growing rate of 0.00075/year. The response of tea NDVI to almost climatic factor in a one-month time lag is higher than the current month. The tea yield was estimated with higher accuracy in the RF model. Among the input variables, we detected that the role of Tmean and NDVI is stronger than other variables when squared with each of the independent variables into input data.

To improve knowledge of this matter, the potential application of two gridded meteorological products (GMPs), the China Meteorological Assimilation Driving Datasets for the SWAT model (CMADS) and Climate Forecast System Reanalysis (CFSR), are compared for the first time with data from ground-based meteorological stations over 6 years, from 2008 to 2013, over the Cau River basin (CRB), northern Vietnam. Statistical indicators and the Soil and Water Assessment Tool (SWAT) model are employed to investigate the hydrological performances of the GMPs against the data of 17 rain gauges distributed across the CRB. The results show that there are strong correlations between the temperature reanalysis products in both CMADS and CFSR and those obtained from the ground-based observations (the correlation coefficients range from 0.92 to 0.97). The CFSR data overestimate precipitation (percentage bias approximately 99%) at both daily and monthly scales, whereas the CMADS product performs better, with obvious differences (compared to the ground-based observations) in high-terrain areas. Regarding the simulated river flows, CFSR-SWAT produced “unsatisfactory”, while CMADS-SWAT (R² > 0.76 and NSE > 0.78) performs better than CFSR-SWAT on the monthly scale. This assessment of the applicative potential of GMPs, especially CMADS, may further provide an additional rapid alternative for water resource research and management in basins with similar hydro-meteorological conditions.