To monitor the space environment and its effects in the low-Earth sun-synchronous orbit of China’s FY-3 satellite, a comprehensive detection technology based on the type Ⅱ loads of the space environment monitor is proposed. In the process of ground development, various technical indicators of the space environment comprehensive detection payload have been calibrated and experimentally verified by different methods such as standard radiation source, equivalent signal source, particle accelerator and standard magnetic field. The results show that the multi-direction full-spectrum particle detection achieves an energy range of 30 keV–300 MeV, with the accuracy of ≤10%. The magnetic field detection realizes the measurement range of −65023–+65023 nT, with the accuracy of ≤0.73 nT. The potential detection realizes the measurement range of −32.4–+23.7 kV, with the sensitivity of ≤10V. The detection of radiation dose realizes the measurement range of 0–3×104 rad (Si), with the sensitivity of ≤8.3 rad (Si). Through comprehensive observation of particle radiation environment, change of in-situ magnetic field vector, radiation dose accumulation and change of satellite surface potential in satellite operation orbit, the space environment monitor provides necessary data support for space activities, satellite design, space science research and space weather early warning and prediction. 

This paper provides quantitative comparison between the text embedding of a text pre-trained model (i.e., RoBERTa) and a multi-modal pre-trained model (i.e., WenLan). Two quantitative comparison methods are proposed, in an embedding space: representing the semantics of a word using the set of ?-nearest words to it and then analyze the semantic changes of the word in the two spaces using the Jaccard similarity of the two sets; forming pairs between each word and its nearest ? words to analyze the relationship. The results show that the multi-modal pre-training brings more semantic changes for more abstract words (e.g., success, love), and the multi-modal pre-trained model can better differentiate antonyms and discover more hypernyms or hyponyms, while text pre-training works better in finding synonyms. Moreover, multi-modal pre-trained model can construct a more extensive associative relationship between words.

The Delft3D model was used to simulate the flooding process on floodplains of the Lower Yellow River, and the flood disaster parameters such as flooding range and flooding depth were statistically calculated from the model. The process of giving the flood risk’s spatial distribution of the plains includes dividing the hazard-vulnerability units and calculating the hazard and vulnerability of each unit according to the topography and flood-control measures. The United Nations risk assessment index system is used to give out the spatial distribution of the flood risk for each inundation unit. The results show that the relationship between flood risk and flood scenarios response across the study area can be divided into three categories. About 50% of the plain are under “simultaneous flooding and risks”, which means the flood risk level generally varies with the flood scenario. 39.7% of the plains own the trait of “low flooding frequency leads to giant risks”, and for the unit of this category basically has the extreme risk even for the 10-year frequency floods. 11.67% of the plain are stable at a relatively low risk situation, and the flood risk is not affected by the magnitude of the flood, which is classified as “stable low risks”. In order to prevent the flood risk of the Lower Yellow River, the flooding unit of “low flooding frequency leads to giant risks” mainly distributed in Gaocun to Aishan section should give a priority to take the action, and the flood situation of once every ten years or even once every five years should be taken as the reference basis for building flood prevention measures.

Based on the 15 m×15 m spatial distribution raster data of 11 types of land use in 2015 and the landscape pattern index, the landscape pattern characteristics of land use in the counties, floodplains and flood detention areas along the wide reaches of the Lower Yellow River, and gradient characteristics of the landscape patterns of land use in the 18 2km-buffer zones on both sides, are compared and analyzed. The results show that the dry farmland is the dominant landscape within the counties along the wide reaches, followed by the rural residential land, and they account for more than 80% of the total area of the counties. The paddy fields and the other categories of waters have a certain scale and the scale of the grassland is small. Influenced by factors such as functional orientation and natural conditions, there are obvious spatial differences in land use landscape patterns in the counties, floodplains and flood detention areas along the wide reaches of the Lower Yellow River. There are large-scale and densely-distributed construction land in the extra-floodplains area of the counties, the landscape fragmentation of which is high and the patch shape is complex. The landscape diversity of the Dongping Lake flood detention area is high, showing a complex ecological landscape with waters as the main body, and in the other two flood detention areas, the scale of woodland and grassland is small, the patch density and the landscape heterogeneity is low. The overall landscape fragmentation in the wide reaches’ floodplains is lower than that in the extra-floodplains area, the landscape diversity of the wandering reaches’ floodplains is relatively high, and the construction land is relatively large. Within 6 km of the two sides of the wide reaches, the paddy fields and the other categories of waters are large in scale and densely distributed, the scale of the urban land is small, and the overall landscape diversity in this region is high and relatively fragmented; the transition region is from about 6 km to 18 km; the landscape patterns tends to be stable over 20 km away.