Datasets of land use change and flood dynamics in the vietnamese mekong delta

This paper compiles the data associated with a research article published in STOTEN [1]. The data set represents figures, tables, and images illustrating the temporal and spatial distribution of land use and flood dynamics from 2000 to 2020 in the Vietnamese Mekong Delta (VMD). The MODIS imageries were freely accessed online via the NASA website [2] and processed to land use and flood maps based on the algorithms by Sakamoto et al. [3,4]. The MODIS products show a high validation with statistical data and radar satellites [1]. The datasets of flood map and land use, therefore, are available to scientists, engineers, and policy-makers in agricultural management associated with flood management in the VMD. They could be used for policy settings, household livelihood assessment as well as other economic analyses for the VMD region due to the change of land use and flooding dynamics.


Specifications
Agriculture, Environmental Science Specific subject area Analyzing land use change and flooding dynamics in the Vietnamese Mekong Delta (VMD) for a period from 20 0 0 to 2020 to monitor holistically the land use change and flooding situation in the region. Type of data Table  Map How data were acquired MODIS imageries were downloaded freely from the NASA website and processed manually to land use and flood maps based on ENVI and GIS programs. Statistical data of rice and aquaculture areas were accessed via the GSO website to be calibrated with MODIS land use data. Data format Analyzed flooding maps (figure, geotiff) Analyzed land use maps (figure, geotiff, excel sheet) Parameters for data collection MODIS products were processed into flooding and land use data as described in Section 2.

Description of data collection
The Moderate-resolution Imaging Spectro-radiometer (MODIS) was selected as the most suitable satellite to serve for this study. Herein, the MODIS products of MOD09A1 (Terra Surface Reflectance 8-Day Global 500m) are applied to process the maps of land use and flood extension. The imageries have a spatial resolution of 500m and were freely downloaded at the NASA website [2] . Statistical data is collected in this study to validate the satellite land use maps as well as to examine the effect of rice production due to land use change. The areas of rice and aquaculture were collected online from 20 0 0 to 2020 at the General Statistics Office of the Vietnam website [5] .

Value of the Data
• We contribute a continuous dataset of land use/land cover (21 maps) and flood dynamics (567 maps) from 20 0 0 to 2020 with detailed spatial data. Hence, they are useful for the VMD agricultural and aquaculture management plan integrated with flood management in the VMD. • Scientists, administrators, decision-makers, and engineers can benefit from these datasets.
• Flooding data can be used/reused for flood risk management in the VMD, i.e., construction dike planning, calibration of flood extension for hydraulic models [6][7] , planning of residential areas, and rice development. Besides, a comparison of flooding management between the countries in the lower Mekong River Basin (i.e., Cambodia and Vietnam) could be evaluated based on the MODIS flooding maps, which cover the territories of Cambodia and Vietnamese Mekong Delta. • The dataset including 567 flooding maps from 20 0 0 to 2020 is a valuable database to setup a flood monitoring website portal for the VMD region. • In addition, the MODIS land use and flooding products can be used for cross-validation with other satellite products with higher spatial resolutions, i.e., Sentinel, Landsat, Copernicus, etc. to assess the appropriateness and accuracy of land use/land cover and flood maps based on remote sensing.
• Last but not least, MODIS land use maps are important for agricultural and aquaculture management plans. They could be a valuable input for (i) evaluating the household livelihoods of local farmers in the last two decades, (ii) economic analyses of cost and benefits, (iii) assessment of agricultural efficiency and rice productivity due to the conversion of rice cropping patterns from double rice to triple rice.

No.
Value Description Color code (R-G-B) 1 1 Water (flooding, lake, ponds) 0 -0 -255 2 2 Mixture 0 -255 -0 3 0 No water (or no data due to out of the study area) Double rice cropping -mainly in the rainy season 255 -255 -0 sensor with the moderate resolution of 500 m. Excluding the mixed-pixel effect in the MODISderived estimates could give good agreement between the obtained spatial patterns of the multiple rice cropping systems and regional land use data [4] . On the other hand, the statistical data books do not have spatial information on land use and land cover, therefore the MODIS-derived land use data contributes to a much better understanding of the spatial distribution of farming systems and annually change of land use patterns in the VMD [1] . The total rice planted area and aquaculture area are annually calculated as: • Total rice planted area = Single rice + (Double rice type 1 + Double rice type 2) x 2 + Triple rice x 3 • Total aquaculture area = Shrimp-rice farming area + Inland aquaculture area For more information on the land use data estimated from MODIS products, please find the excel sheet "Land use data.xlsx" in the Supplementary Materials.
Besides, Fig. 2 illustrates the maximum flooding extension in the VMD. The flooding maps contain three main classes with water (blue), mixed pixel (green), and non-flooded areas (white). Here, the maximum flooding extension is determined as the date with the largest flooding extent in the VMD, which occurs annually between October 8th and November 9th [1] .
Overall, the analyzed data of flooding and land use are published online for further assessment. The data can be downloaded at https://data.mendeley.com/datasets/kpftzmsyyz/2 . Where: • Flooding data are displayed in GeoTiff format in "Flooding" folder and named "Flood_year_DOY". Here, DOY indicates Day of the Year, and the flooding maps reach annually from early June (Flood_year_153) to end of December (Flood_year_361). The definition of flooding maps is presented in Table 3 . • Land use data set are GeoTiff files in "Land use" folder and named "Land use_year". In which, "year" indicates from 20 0 0 to 2020. The definition of land use maps is displayed in Table 4 .

Experimental Design, Materials and Methods
The imageries of MODIS products were processed for flood mapping and land use detection in the VMD according to algorithms developed by Sakamoto et al. [3] and Sakamoto et al. [4] . The approaches of flood mapping and land use detection apply three important indexes, i.e., Enhanced Vegetation Index (EVI), Land Surface Water Index (LSWI), and Difference Value between EVI and LSWI (DVEL), which are expressed as: where RED is the red band (sur_refl_b01), NIR is the near-infrared band (sur_refl_b02), SIWR is the short-wave infrared band (sur_refl_b06), and BLUE is the blue band (sur_refl_b03) of the MODIS surface reflectance. The flood maps were processed into three classes including (i) Flood; (ii) Mixed pixel; and (iii) No water (see Table 3 ). While the output of the land use maps contains eleven classes Vu et al. [1] . In this paper, we grouped the classes of land use maps into seven objects to be more convenient for users, see Table 4 . Hence, the land use maps include shrimp-rice farming, inland aquaculture, single rice, double rice cropping in the dry season, double rice cropping in the rainy season, triple rice cropping, and others (forest, orchard, extensive farming, mixtures in flood-prone areas, mixtures in double rice and triple rice cropping, and unused areas).
A detailed description of flood and land use processing could be referred to Vu et al. [1] , Sakamoto et al [3] ; Sakamoto et al. [4] , and/or Vu [8] .

Ethics Statement
This work does not involve chemicals, procedures or equipment that have any unusual hazards inherent in their use. This work does not involve the use of animal or human subjects.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships which have or could be perceived to have influenced the work reported in this article.

Data Availability
Datasets of Land Use Change and Flood Dynamics in the Vietnamese Mekong Delta (Original data) (Mendeley Data).