Month: January 2025
The Separation Crosses the Strait
Abstract
During the 1990s, the distinction between mainlander engineers and local engineers appeared to blur as both participated enthusiastically in extending Taiwan-owned companies with Taiwan-developed technologies across the Taiwan Strait to mainland China. Examined in this chapter, enrollments in engineering degree programs at Taiwanese universities basically exploded as the KMT government elevated junior colleges to the status of universities. A preponderance of mostly local students pursued degree programs in electronics engineering and computer science. Also, over the next decade large numbers of native Taiwanese returned from working abroad, finally overcoming the sense that career opportunities and family responsibilities had to pull one in opposite directions. By establishing and maintaining a physical presence on the mainland, mainland Taiwanese engineers and local Taiwanese engineers carried with them distinct images of Taiwan’s relationship with China. The distinction between them was not blurred as they effectively traveled across the strait on two pathways. One pathway sought to maintain centeredness and control in Taiwan while the other aimed to build productive power with mainland China. The manufacture of semiconductors became a key site of tension between them. Then engineers from the mainland itself enacted yet another geographical identity as the Chinese government increasingly asserted its own vision of the meanings and implications of cross-strait relations.
To the Engineers of Taiwan, Past and Present
Abstract
Taken together, the five formative episodes demonstrate that congruences between the technical identities and geographical identities of engineers have both granted them privileges and assigned them responsibilities as they helped to make and re-make the land of Taiwan. Chapter 7 isolates the specific relationships that emerged between engineers and the island through infrastructures of education/training and work, showing how those privileges and responsibilities extend through the present.
Abnormal Engineers for an Abnormal Land
Abstract
Our central method has been to trace the development of infrastructures of engineering formation (including education and training) and engineering work across Taiwan, looking for key episodes in the changing relationships between emergent engineers and the island. Infrastructures of engineering formation include physical, organizational, and abstract structures ranging from school buildings to government ministries to course curricula and contents. Infrastructures of engineering work range from actual work practices to organizational roles on the job and in professional societies to the positioning of engineering work in relation to larger missions. We explore five formative episodes: the Japanese colonial period, early KMT period, development of the electronics industry, emergence of professional societies in the United States, and, finally, a period characterized by the flows of engineers from Taiwan to the mainland. Infrastructures of formation and work in these episodes enabled particular images of engineers and engineering to emerge, travel, and become dominant for those who faced them. Also, prospective and working engineers who passed through such infrastructures or otherwise encountered them were challenged to accept forms of knowledge that pointed their work and careers in particular directions, with associated identities and commitments. Through a selective collection and marshalling of relevant evidence, the chapters show together how congruences between technical and geographical identities had the effect of attaching emergent engineers to the land of Taiwan as it evolved to become a thing, an entity, unto itself. The level of detail in the episodes might suggest that they are histories. They are not. They are sharply delimited investigations of relationships between the pathways of engineers into and through engineering and their connections to the land of Taiwan.
Freelancing Around Japanese Infrastructures
Abstract
The episodes in Chaps. 2 and 3 explore the struggles of would-be local engineers and engineering technicians to climb industrial infrastructures that were controlled by others. During the 50-year colonial period examined in Chapter 2, the Japanese empire understood its engineers as technical talents who had knowledge of machinery, and engineers on the island nearly always came from the Japanese mainland. Most local boys and men that went to work in machine-based industry were educated only in common schools and had to join as manual laborers. For the most part, those who were able to gain access to some machine knowledge in new industrial schools had to work as assistants to Japanese technicians or, later, travel to Manchukuo. Yet some were able to avoid direct Japanese supervision and become “freelance” workers, akin to doctors, who established small companies that hired local labor and built spaces to operate through contractual relationships with Japanese companies. And some freelancers who worked successfully outside of Japanese companies began to embrace the land of Taiwan apart from the empire by claiming the identity of professionals, a category of person that is typically seen as detached from geography. As Taiwan became a base for southern expansion and key contributor to the Greater East Asia War, many local companies flourished. And by 1945, local industrial technicians and would-be engineers were contributing significantly not only alongside Japanese companies but also within them.
Unsupervised Meteorological Downscaling Based on Dual Learning and Subgrid-scale Auxiliary Information
Abstract
Climate downscaling is used to transform large-scale meteorological data into small-scale data with enhanced detail, which finds wide applications in climate modeling, numerical weather forecasting, and renewable energy. Although deep-learning-based downscaling methods effectively capture the complex nonlinear mapping between meteorological data of varying scales, the supervised deep-learning-based downscaling methods suffer from insufficient high-resolution data in practice, and unsupervised methods struggle with accurately inferring small-scale specifics from limited large-scale inputs due to small-scale uncertainty. This article presents DualDS, a dual-learning framework utilizing a Generative Adversarial Network–based neural network and subgrid-scale auxiliary information for climate downscaling. Such a learning method is unified in a two-stream framework through up- and downsamplers, where the downsampler is used to simulate the information loss process during the upscaling, and the upsampler is used to reconstruct lost details and correct errors incurred during the upscaling. This dual learning strategy can eliminate the dependence on high-resolution ground truth data in the training process and refine the downscaling results by constraining the mapping process. Experimental findings demonstrate that DualDS is comparable to several state-of-the-art deep learning downscaling approaches, both qualitatively and quantitatively. Specifically, for a single surface-temperature data downscaling task, our method is comparable with other unsupervised algorithms with the same dataset, and we can achieve a 0.469 dB higher peak signal-to-noise ratio, 0.017 higher structural similarity, 0.08 lower RMSE, and the best correlation coefficient. In summary, this paper presents a novel approach to addressing small-scale uncertainty issues in unsupervised downscaling processes.
Convolutional Graph Neural Network with Novel Loss Strategies for Daily Temperature and Precipitation Statistical Downscaling over South China
Abstract
Traditional meteorological downscaling methods face limitations due to the complex distribution of meteorological variables, which can lead to unstable forecasting results, especially in extreme scenarios. To overcome this issue, we propose a convolutional graph neural network (CGNN) model, which we enhance with multilayer feature fusion and a squeeze-and-excitation block. Additionally, we introduce a spatially balanced mean squared error (SBMSE) loss function to address the imbalanced distribution and spatial variability of meteorological variables. The CGNN is capable of extracting essential spatial features and aggregating them from a global perspective, thereby improving the accuracy of prediction and enhancing the model’s generalization ability. Based on the experimental results, CGNN has certain advantages in terms of bias distribution, exhibiting a smaller variance. When it comes to precipitation, both UNet and AE also demonstrate relatively small biases. As for temperature, AE and CNNdense perform outstandingly during the winter. The time correlation coefficients show an improvement of at least 10% at daily and monthly scales for both temperature and precipitation. Furthermore, the SBMSE loss function displays an advantage over existing loss functions in predicting the 98th percentile and identifying areas where extreme events occur. However, the SBMSE tends to overestimate the distribution of extreme precipitation, which may be due to the theoretical assumptions about the posterior distribution of data that partially limit the effectiveness of the loss function. In future work, we will further optimize the SBMSE to improve prediction accuracy.