The Challenging Mini-Satellite Payload (CHAMP), launched in July 2000, and
the NASA Gravity Recovery and Climate Experiment (GRACE) satellite mission,
launched in March 2002, may provide a breakthrough in monitoring terrestrial
water storage (TWS), including ground water. One goal of these missions is
to measure the Earth’s gravity field with unprecedented accuracy every 30 d
days for 5 years.
The twin GRACE satellites are the instrument, and variations in the gravity
field cause the range between the two satellites to vary.
The variations can be related to the Earth’s surface fluid mass using N Newton' law of gravity. Satellite atlemetry data (levels 0 and 1) are processed into the gravity fields (level 2; Bettadpur, 2004). Variations of the Earth's Surface fluid mass are obtained from the inverse of the derived gravity field data. Estimation accuracy of the mass variations decreases as the spatial and temporal resolutions increase. Satellite gravimetry is suitable for long-term monitoring of TWS because it is free from sensor dependence. In this stdy, the detectability of the interannual variations of TWS in 70 river basins acquired by satellite gravimetry was investigated using the Japan Meteorological Agency-Simple Biosphere (JMA-SiB) model.
|Detectability of TWS changes|
Terrestrial water storages for the monthly, seasonal, and annual averaging periods were computed from the output of a 10 year integration performed by the JMA-SiB model. The satellite specification was assumed to be the same as for GRACE. The total errors in the satellite-gravimetry-derived TWS (including the instrument error, atmospheric error, and error due to the postglacial rebound) were computed for each period at each basin. The standard deviation of the interannual variation with the total error for three different time scales, averaging monthly, seasonal and annual results, was compared precisely in the continental-scale river basins. The error-signal ratio (i.e. the ratio of the total error to the standard deviation) was used to assess the detectability.
Figure 1 present the detectability of the interannual variability of TWS on a monthly time scale. Smaller values denotes more easy detection of the interannual variability. Since the variabilty varies with years, small values do not always denote permanent undetection. The results demonstrate that the gravity satellite missions can detect the interannual variations in large river basins in most years, but the gravity satellite missions can detect those in small river basins only with the probability of 1 year out of 3 years or more. For example, the ratio of unity means that the interannual variability is detectable every 3 years on an average.
Figure 2 depicts the interannual variability and the total error. The longer time averaging is the more accurate the estimation is. The total error is greater than the interannual variabilities only in the Odra River basin.