Ground and satellite based observation datasets for the Lower Mekong River Basin

In ‘Satellite observations and modeling to understand the Lower Mekong River Basin streamflow variability’ [1] hydrological fluxes, meteorological variables, land cover land use maps, and soil characteristics and parameters data were compiled and processed for the Lower Mekong River Basin. In this work, daily streamflow time series data at nine gauges located at five different countries in the Mekong region (Thailand, Laos People׳s Democratic Republic (PDR), Myanmar, Cambodia, and Viet Nam) is presented. Satellite-based daily precipitation and air temperature (minimum & maximum) data is processed and provided over the entire basin as part of the dataset provided in this work. Moreover, land cover land use raster data that contains 18 classes that cover agriculture, urban, range and forests land cover land use classes for the basin is offered. In addition, a soil data that contains physical and chemical characteristics needed by physically based hydrological models to simulate the cycling of water and air is also provided.


a b s t r a c t
In 'Satellite observations and modeling to understand the Lower Mekong River Basin streamflow variability' [1] hydrological fluxes, meteorological variables, land cover land use maps, and soil characteristics and parameters data were compiled and processed for the Lower Mekong River Basin. In this work, daily streamflow time series data at nine gauges located at five different countries in the Mekong region (Thailand, Laos People's Democratic Republic (PDR), Myanmar, Cambodia, and Viet Nam) is presented. Satellitebased daily precipitation and air temperature (minimum & maximum) data is processed and provided over the entire basin as part of the dataset provided in this work. Moreover, land cover land use raster data that contains 18 classes that cover agriculture, urban, range and forests land cover land use classes for the basin is offered. In addition, a soil data that contains physical and chemical Value of the data The satellite dataset benefits hydrologic modeling at poor spatial in-situ data regions such as the Lower Mekong River Basin.
The dataset can assist to understand the water balance in the Lower Mekong River Basin. The dataset is essential in hydrological modeling in the Mekong region since it contains new developed land cover land use, and soil characteristics layers.

Data
This paper reports various hydrological time series and remote sensing data that was used to model and understand the streamflow variability in the Lower Mekong River Basin (LMRB) [1]. Mohammed et al., [1] used the Soil & Water Assessment Tool (SWAT) hydrologic model (https://swat.tamu.edu/) to simulate hydrological fluxes in the LMRB and explore the streamflow regime changes as a result of expected upstream flow changes (i.e., the Chinese part of the Mekong River). The LMRB streamflow data reported is observed and collected by various agencies in the Lower Mekong Region. The LMRB streamflow time series data that we present here was obtained from the Mekong River Commission (MRC) hydrological respiratory (http://www.mrcmekong.org/). Interpolation was carried out using a recent observed level data acquired from the Asian Preparedness Disaster Center (ADPC) geospatial and climate resilience teams (http://www.adpc.net/) to update and fill the gaps of the streamflow time series data we present. The dam data we report here was obtained from the Greater Mekong Consultative Group for International Agricultural Research (CGIAR) Program on Water, Land and Ecosystems [2].
The precipitation data we report here was obtained from the Tropical Rainfall Measurement Mission (TRMM) [3], and combined with the Integrated Multi-satellite Retrieval for the Global Precipitation Measurement mission (IMERG) [4] remote sensing data products. The TRMM, and GPM remote sensing data products can be accessed at https://pmm.nasa.gov/data-access/downloads/. The TRMM dataset we report here was processed from a daily 0.25 Â 0.25°accumulated precipitation that is generated from the near real-time 3-hourly (TMPA /3B42RT) product. We also report precipitation data obtained from the IMERG dataset. The IMERG dataset presented here is the Global  Table 1. The dams shown are outlined in Table 2.
Precipitation Mission (GPM) Level 3 IMERG *Final* Daily 0.1 Â 0.1 degree (GPM_3IMERGDF) data product, which is derived from the half-hourly data product (GPM_3IMERGHH). The derived result represents the final estimate of the daily accumulated precipitation in millimeters.
Minimum and maximum daily air temperature data we describe here was calculated from air temperature record retrieved from the Global Land Data Assimilation System (GLDAS) simulation data products [5]. The goal of the GLDAS [5] is to ingest satellite and ground-based observational data products, using advanced land surface modeling and data assimilation techniques, in order to generate optimal fields of land surface states and fluxes. For this paper, we used the GLDAS Noah Land Surface Model L4 3-h 0.25 Â 0.25°(GLDAS_NOAH025_3H.2.1) data product available at https://disc. gsfc.nasa.gov/.
The land cover land use data we present was produced from multiple 2010 land cover land use maps at a spatial resolution of a 0.25 km for the Lower Mekong Basin [6]. The land cover land use maps produced and presented herein used the Moderate Resolution Imaging Spectroradiometer (MODIS) monthly Normalized Difference Vegetation Index (NDVI) data and circa 2010 dry season Landsat reflectance data as the primary data sources as well as high resolution satellite data from Google Map/Earth, and other reference data from the MRC. The spatial scale of the land cover land use data presented here is 90 m.
The soil data we report here was processed from the Harmonized World Soil Database (HWSD) [7]. The HWSD data was obtained from the Food and Agriculture Organization of the United Nations (FAO) and processed to be compatible with the SWAT hydrological model. The HWSD soil data is a 30 arcsecond raster database with over 15,000 different soil mapping units that combines existing regional and national updates of soil information worldwide.

Experimental design, materials, and methods
The streamflow time series data is presented with this paper as a spreadsheet in Appendix A.1. The streamflow gauge names, gauge identification codes, and record period for each gauge is shown in Table 1.
Dams within the Lower Mekong River Basin that are either already commissioned or still under construction and have a maximum reservoir area greater than or equal to 280 km 2 are reported in this paper. Table 2 gives dam data at the LMRB that was used in simulating streamflow [1].
The precipitation data for the whole LMRB is presented with this paper as a spreadsheet named 'Precipitation' in Appendix A.2 which gives climate data for the study area. The precipitation data units are in millimeters. The temporal span for the data is from 1 January 2001 to 31 December 2015. Area weighted average methodology was used to obtain an aggregated precipitation time series data for the LMRB. Since IMERG data products are only available from 12 March 2014 to present, we used the TRMM rainfall data (3B42RT) for time periods earlier than 12 March 2014. A nearest neighbor methodology was used to fill the IMERG data points with the TRMM data points as an approximation during the 1 January 2000 to 11 March 2014 time period [8]. Since TRMM and IMERG data do not have  n/a n/a n/a n/a n/a n/a n/a Data provided by IWRP the same spatial resolution (i.e., 0.25 and 0.1 degree respectively), a methodology was presented in Mohammed et al. [8] to address the spatial scale differences. The air temperature data for the whole LMRB is presented with this paper as spreadsheets named 'Tmin' and 'Tmax' in Appendix A.2. The air temperature data (minimum and maximum) units are in degree Celsius. We calculated the daily minimum and maximum temperatures by finding the minimum and maximum air temperatures for each day at each grid within the study watershed by searching for minima and maxima over the three hours air temperature data values available for each day and grid. Area weighted average methodology was used to obtain an aggregated air temperature (min/max) time series data for the LMRB.
The MODIS monthly NDVI images used to produce the land cover land use map presented (Appendix A.3) were derived from MOD09 and MYD09 8-day reflectance data that was temporally processed using the Time Series Product Tool custom software package [9]. The land cover land use map produced were developed primarily from unsupervised classification of the 2010 MODIS NDVI data, including several agricultural and forest types. The Landsat data was used with a combination of unsupervised and supervised classification methods to map land cover land use classes that were regionally scarce but locally common, including bamboo forest scrub, industrial forest plantation, urban, and water classes. Geographic Information System techniques were then applied to integrate the MODIS and Landsat classifications into singular land cover land use map for entire LMRB. The land cover land use classes presented here for the LMRB are listed in Table 3. In general, the land cover land use classes can be categorized into agricultural land classes, forest type classes, grass lands, urban lands, and water. Appendix A.3 gives the land cover land use raster grid for the LMRB along with the raster projection information.
The soil data reported here was produced to meet the hydrological modeling needs in Mohammed et al., [1]. The soil database development was intended to be as an input for the SWAT model development at the Lower Mekong. However, the methodology and parameters presented here can aid other studies. The saturated hydraulic conductivity, a quantitative measure of a saturated soil's ability to transmit water when subjected to a hydraulic gradient, reported here as "SOL_K(layer#)" has been estimated using the Soil Water Characteristics and the SPAW Hydrology and Water Budgeting software [10]. Table 4. gives a summary of the various soil parameters presented in the Appendix A.4 soil table 'LMRB_usersoil.xlsx'. Appendix A.4 also gives the soil raster grid for the LMRB using the same projection highlighted earlier in the land cover land use raster grid (Appendix A.3). Table 3 Land cover land use classifications. Raster value refers to the land cover land use ascii raster file provided in Appendix A.3.