*4.2. Soil Dryness Response to Large Rainfall Events*

In this section, soil moisture is examined alongside precipitation with the purpose of examining the drying of soil over time in response to a rainfall event. By evaluating the time series after a large precipitation event with almost no subsequent precipitation, we were able to observe the near-surface soil moisture observations as they transitioned from saturated to dry conditions. Daily 9 km SMAP soil moisture estimates were compared to daily 10 km IMERG precipitation to examine the response of soil moisture to precipitation events. It is possible that, in the absence of precipitation, agriculture is irrigated. Hence, we may have seen wetness from irrigation in these regions, despite no significant rainfall event. Figure 5 shows the relationship between daily rainfall and soil moisture between 2015 and 2018 averaged over the LMB.

**Figure 5.** Time series of daily soil moisture active passive (SMAP) 9 km soil moisture and daily Global Precipitation Measurement-Integrated Multi-satellitE Retrieval (GPM-IMERG) 10 km precipitation for April 2015–September 2018 averaged in the LMB.

Using Figure 5, two precipitation events were selected in which soil moisture exhibited a clear dry-down pattern after the rainfall. The events were examined more closely in combination with corresponding daily downscaled soil moisture, in order to evaluate the improvement in the representation of drying from 9 km to 1 km. Figure 6 more closely examines the time series of the dry-down period in the LMB from 13 April 2015 to 20 April 2015, after a large precipitation event occurred on 13 April. Figure 7 shows the spatial distribution of rainfall, 9 km SMAP soil moisture, and 1 km downscaled soil moisture for each day during the dry-down period. The second event selected was from 6 April 2018 to 11 April 2018. Figure 8 shows the time series of the dry-down period after the precipitation event on 6 April 2018. The 1 km soil moisture (blue) was better able to capture the dry-down pattern than the 9 km SMAP soil moisture (green) (Figure 8). Figure 9 shows the spatial distribution of rainfall, 9 km SMAP soil moisture, and 1 km downscaled soil moisture for each day during the dry-down period in April 2018. The coverage of the 1 km corrected soil moisture was dependent on MODIS LST data and influenced by cloud cover, which made it difficult to find good coverage on consecutive days. The 1 km SMAP did not perform as well during wet days due to the spatial coverage of the MODIS land surface temperature (LST) data being compromised by cloud contamination.

**Figure 6.** Time series of dry-down event from 13 April 2015 to 20 April 2015 with 9 km SMAP soil moisture (green), 1 km downscaled soil moisture (blue), and 10 km rainfall data from GPM IMERG (black).

**Figure 7.** *Cont*.

**Figure 7.** Dry-down event for 13 April 2015 to 20 April 2018 represented by 10 km IMERG rainfall, 9 km SMAP, and 1 km downscaled soil moisture.

**Figure 8.** Time series of dry-down event from 6 April 2018 to 11 April 2018 with 9 km SMAP soil moisture (green), 1 km downscaled soil moisture (blue), and 10 km rainfall data from GPM IMERG (black).

**Figure 9.** Dry-down event for 6 April 2018 to 11 April 2018 represented by 10 km IMERG rainfall, 9 km SMAP, and 1 km downscaled soil moisture.

#### *4.3. Importance of High Spatial Resolution Soil Moisture for Hydrology and Water Resources*

The high spatial resolution observed soil moisture generated in this study was an important data set that could not be obtained from other sources. Firstly, there are no consistent in situ networks that monitor soil moisture in the Lower Mekong River Basin. Even in other parts of the world that do have such networks, they are seldom dense enough to produce soil moisture at 1 km spatial resolution. Secondly, although land surface models can simulate soil moisture at high spatial resolution, they lack the precipitation input at 1 km spatial resolution, which is needed to minimize variations in small-scale processes [35]. Currently, the "best" spatial resolution of globally available precipitation is the climate hazards group infrared precipitation with station observations (CHIRPS) at 0.05◦. CHIRPS provides estimates from 1981 to the near present and uses a recently produced satellite rainfall algorithm that combines climatology data, satellite precipitation estimates, and in situ rain gauge measurements to produce a high resolution precipitation product [36].

The 1 km spatial resolution soil moisture from this research can be used in combination with land use and land cover data from MODIS (moderate resolution imaging spectroradiometer) at 1 km and Landsat imagery at 30 m to map the co-variability of land use and wetness. This will be a valuable tool for land use planning, specifically in the LMB where there are competing cropping strategies and land use for industrial development. Additionally, this 1 km soil moisture can be used to determine antecedent soil moisture conditions in watershed modeling, meaning it can serve as an input to determine the portion of rainfall that will infiltrate the soil and that which will run off to the stream network. More detailed estimations of streamflow runoff will in turn benefit flood prediction and monitoring in watersheds [37]. This high spatial resolution 1 km observed soil moisture can serve a variety of water resource applications and will be of much use in the LMB.
