
Version Summary
- Processing data for all global land regions. These data are considered experimental, and no quality checks are being performed for the data outside of the polar regions.
- Dynamically computing the radial component of geolocation uncertainty (sigma_geo_r). Previously, this value was fixed. The uncertainty is calculated for ATL03 (sigma_h) and the value is passed through to ATL06 (sigma_geo_r). Values for the horizontal geolocation uncertainties (sigma_geo_at and sigma_geo_xt) remain at fixed, pessimistic values of 5 m.
Overview
Strengths
- This data set is suitable for assessing the change of land ice height (especially ice sheets) (Shen et al., 2020) at a seasonal scale or longer because the ICESat-2 satellite has a repeat cycle of 91 days and the file names contain convenient spatio-temporal information (cycle number and ground track).
- This data set provides geolocated land ice surface heights with centimeter-scale precision and several-meter horizontal positioning accuracy (Brunt et al., 2020) along 40 m segments of ground track, spaced 20 m apart (Smith et al., 2019). Brunt et al. (2020) assessed the accuracy of ATL06 mostly over relatively flat terrain (i.e., ice sheets) and have concluded that it is more accurate and precise than any prior altimetric product.
- Measurement bias has been well-characterized, and strategies used to mitigate biases are outlined in the User Guide (ATL06 User Guide; ATL06 ATBD). Corrections to account for these biases are either applied to surface height or are included as an accompanying field, contributing to ATL06’s improved accuracy and precision over ATL03 (Smith et al., 2020).
Limitations
- For the assessment of land ice mass change, height measurements provided by this product must be corrected for changes in the density of the firn (firn air content) and corrections for solid earth vertical motion through time. Estimates of firn air content are currently associated with large uncertainties (Smith et al., 2020).
- The accuracy of these data is not regionally consistent and depends on local surface conditions and terrain (Shen et al., 2020). Data assessments report that errors increase with increasing surface roughness and slope (Brunt et al., 2019; Shen et al., 2020), so height information is less accurate in areas with high surface roughness and complex terrain features (i.e., slopes greater than 20 degrees or in proximity to a high degree of crevassing).
- The ATL06 product is not corrected for the atmospheric and subsurface scattering that delays photon travel-time, resulting in negatively biased surface measurements (ATL06 ATBD, ATL06 User Guide). Studies have reported a small systematic low bias (Shen et al., 2020) and noted that the loss of accuracy due to atmospheric scattering is affected most significantly when cloud optical thickness exceeds a value of 2 (in the case of smooth surfaces) (Smith et al., 2019).
- ATL06 is derived from the ATL03 product, which does not always correctly identify signal ground-return photon events. In some cases, clusters of misidentified photons can result in relatively larger errors (ATL06 User Guide).
- Since ATL06 reports elevations in 40 m overlapping segments, it cannot resolve features such as cracks and rifts that are smaller than this (Herzfeld et al., 2020).
- In rough or steeply sloping areas of the ice sheet, quality of retrieved segments may be low, as indicated by the atl06_quality_summary flag; however, some flagged segments may be of usable quality if they can be reliably identified with another method.