4:15 PM - 4:30 PM
[ACG37-10] Estimation of non-photosynthetic vegetation amount in the northern Gobi Desert using MODIS satellite data
Keywords:non-photosynthetic vegetation , Soil Tillage Index, roughness length, threshold friction velocity, arid regions
The focus of monitoring vegetation amount has been mainly on green vegetation during the growing season. Comparatively less research has been done in estimations of non-photosynthetic vegetation (NPV) that includes litter and standing dead vegetation, although monitoring its coverage and mass is of great importance in various applications in arid and semi-arid regions, such as fire risk management, livestock forage assessment and wind erosion risk assessment. The present study intended to estimate the NPV amount based on MODIS vegetation indices and field observation data at one of the main source regions for Asia Dust, Tsogt-Ovoo (TsO) located in the northern Gobi Desert, Mongolia.
We compared the measured NPV coverage (fcNPV) and mass (mNPV) with NDVI and other six potential vegetation indices using the short wavelength infrared region (SWIR) bands. We found that the combination of band 6 (1.6 µm) and band 7 (2.1 µm), namely the Soil Tillage Index (STI) and Normalized Difference Tillage Index (NDTI), had a good performance in estimating fcNPV and mNPV.
We propose the roughness length z0 and threshold friction velocity (u*t) for sand saltation, which had been observed at non-stony and stony sites in TsO, as representative parameters for the vegetation amount. Using the parameters, we verified suitability of fcNPV estimated from STI. At both sites, the observed z0 and u*t increased as the estimated fcNPV increased. We established the relationship between the observed z0 and the estimated fcNPV at the sites. The predicted z0 in the absence of NPV were 0.0010 cm and 0.0085 cm at non-stony and stony sites, respectively. These values are reasonable in accordance with those of other studies. We also established the relationship between the observed u*t and the estimated fcNPV. The predicted u*t in the absence of NPV were 0.23 m s-1 and 0.30 m s-1 at non-stony and stony sites, respectively. The threshold wind speed at 10 m height (u10t0) converted from u*t were 5.3 and 6.7 m s-1, close to the value of 6.5 m s-1 that commonly used in dust models.
Those verifications indicate that STI is suited to estimate the NPV amount at least at TsO in the Gobi Desert.
We compared the measured NPV coverage (fcNPV) and mass (mNPV) with NDVI and other six potential vegetation indices using the short wavelength infrared region (SWIR) bands. We found that the combination of band 6 (1.6 µm) and band 7 (2.1 µm), namely the Soil Tillage Index (STI) and Normalized Difference Tillage Index (NDTI), had a good performance in estimating fcNPV and mNPV.
We propose the roughness length z0 and threshold friction velocity (u*t) for sand saltation, which had been observed at non-stony and stony sites in TsO, as representative parameters for the vegetation amount. Using the parameters, we verified suitability of fcNPV estimated from STI. At both sites, the observed z0 and u*t increased as the estimated fcNPV increased. We established the relationship between the observed z0 and the estimated fcNPV at the sites. The predicted z0 in the absence of NPV were 0.0010 cm and 0.0085 cm at non-stony and stony sites, respectively. These values are reasonable in accordance with those of other studies. We also established the relationship between the observed u*t and the estimated fcNPV. The predicted u*t in the absence of NPV were 0.23 m s-1 and 0.30 m s-1 at non-stony and stony sites, respectively. The threshold wind speed at 10 m height (u10t0) converted from u*t were 5.3 and 6.7 m s-1, close to the value of 6.5 m s-1 that commonly used in dust models.
Those verifications indicate that STI is suited to estimate the NPV amount at least at TsO in the Gobi Desert.