Comparison of active and passive water vapor remote sensing from space: An analysis based on the simulated performance of IASI and space borne differential absorption lidar

Abstract

Two remote sensing systems, which are considered to be operated in space, the Infrared Atmospheric Sounding Interferometer (IASI) and the Water Vapour Lidar Experiment in Space (WALES) are compared with respect to their measurement methodologies and their performance. The focus is the retrieval of water vapor, which is determined by the inversion of the radiative transfer equation in case of IASI and by the differential absorption lidar (DIAL) technique in case of WALES. It is realized that different techniques and definitions for the specification of errors exist which are subject of confusion in the remote sensing community. After a clarification of this issue, a comparison of IASI and WALES water vapor retrievals is performed using the same water vapor climatologies and the same 2-d water vapor fields provided by a global numerical weather prediction (NWP) model. The methodologies and the capabilities of each instrument are compared in regions without clouds. Using end-to-end simulations for both instruments, which are to our knowledge performed for the first time, systematic errors are compared up to 16 km. It is found that the dependence of IASI retrievals on a variety of atmospheric parameters leads to compensating effects. Due to the multiwavelength retrieval, errors in the water vapor spectroscopy can partly cancel. The residual error is quantified by inversion of the radiative transfer equation in dependence of several atmospheric variables. In contrast, errors in water vapor DIAL are very sensitive to laser spectral properties as well as to the accuracy of water vapor spectroscopy, as single water vapor absorption lines are used for each vertical segment of the retrieval. As laser transmitters with excellent spectral specifications are feasible, this can still lead to very low systematic errors under all atmospheric conditions. Noise errors are determined using analytical models and are compared up to 16 km. At the same vertical (1–2 km) and horizontal (100–200 km) resolutions, respectively, the average noise errors in each profile are of the order of 10% for both methods. Depending on the climatology, the vertical range of IASI measurements is always several kilometers lower than that of DIAL. The performance of IASI degrades in dry atmospheres whereas the DIAL performance remains nearly independent of the climatology chosen. Bias errors show a similar behavior. Neglecting bias errors in the spectral measurements, from mid-latitudes to the tropics, IASI biases are <2 % in the vertical range where the noise errors remain <20%. In the sub-arctic winter atmosphere, the bias increases to about −4% close to the ground. Space borne DIAL bias profiles range between −2–1% under all conditions plus an additional height independent bias of about ±2 due to remaining uncertainties in absorption line spectroscopy. Operation in the region of clouds are not a focus of this publication but it is worth to mention that the results demonstrate that space borne DIAL can perform measurements down to cloud tops and often through optically thin clouds. Particularly powerful is the synergistic combinations of both sensors in the future. Iteration between IASI temperature and DIAL water vapor retrievals will increase both accuracies.