Conference Agenda

The Third Pole Environment plays a significant role in global atmospheric circulation and is highly sensitive to climate change and its impact on Asia’s largest rivers which provide water to 1.5 billion people across ten countries. A fundamental understanding of intensive exchanges of water and energy fluxes between the Asian monsoon, the plateau land surface and the plateau atmosphere at various temporal and spatial scales especially at the climate scale is crucial to understand the role of TPE on global climate and the impact of climate change on TPE.

The CLIMATE-TPE project aims to advance the understanding of the interactions between the Asian monsoon, the plateau surface (including its permafrost and lakes) and the Tibetan plateau atmosphere in terms of water and energy budgets in order to assess and understand the causes of changes in cryosphere and hydrosphere, in relation to changes of plateau atmosphere in the Asian monsoon system, and to predict the possible changes in water resources in the Third Pole Environment. A core innovation of the project is to verify or falsify recent hypotheses (e.g. links between plateau heating and monsoon circulation, snow cover and monsoon strength, soil moisture and timing of monsoon) and projections of the changes of glaciers and permafrost in relation to surface and tropospheric heatings on the Tibetan plateau as precursors of monsoon pattern changes and glaciers retreat, and their impacts on water resources in South East Asia.

We use earth observation, in-situ measurements and modelling to advance process understanding relevant to monsoon scale predictions, and improve and develop coupled regional scale hydroclimatic models to explain different physical links and scenarios that cannot be observed directly. Three work-packages (WP) are defined to address three specific objectives. 1) advancement of the understanding of microwave scattering and emission under complex terrains with permafrost and freeze – thawing conditions. The focus is to reduce uncertainties in current microwave satellite observations over complex terrain and improve retrieval accuracies of soil moisture and freeze-thaw states by deploying in-situ observations, laboratory experiment and numerical modelling. 2) Advancement of physical understanding and quantification of changes of water and energy budgets in the TPE. The focus here is to integrate current understandings in the mechanism of changes in water and energy budget in TPE using satellite data products and numerical modelling. Objective 3) Advancement of quantifying changes in surface characteristics and monsoon interactions. All variables related to water and energy budgets in TPE will be subject to systematic analysis to ensure their consistence in terms of climate data records. The variables will include albedo, vegetation coverage, soil thermal and hydraulic properties, LST, soil moisture, lake levels and land use changes among others.

In this contribution we focus on WP1: Observation and modelling of microwave scattering and emission under complex terrains and including permafrost and freeze and thawing; and WP2: Advancement of physical understanding and quantification of changes of water and energy budgets in the TPE.

Since 2006 the Tibetan plateau observatory for soil moisture and soil temperature (Tibet-Obs, Su et al., 2011, HESS) has been in operation and has provided valuable dataset for land-atmosphere process studies. The networks and collected data have been used for calibration and validation of satellite soil moisture retrieval algorithms and data products as well as for improving numerical model parameterizations (Su et al., 2013, JGR; Zheng et al., 2015a, b, JHM; 2017a, JHM, b, JGR) and for understanding passive and active microwave signals (Dente et al., 2015, RSE; Wang et al., 2016, JAG; Lv et al., 2014, RSE). Most recently an in-situ microwave radiometer (ELBARA III from ESA) has been operating at the Maqu site of the Tibet-Obs, as such coherent process observation, process modeling and radiative transfer modeling can be conducted (Zheng et al., 2017, TGRS) to examine land-atmosphere interactions. We report here recent results of these experiments in combined radiative transfer and heat-water transfer processes and in understanding satellite observation signals and data products – these are related to a new insight of the penetration depth and its quantification for soil moisture products (Lv et al., 2018, RS; Lv et al., 2019, TGRS), benefit of synergistic use of active and passive microwave observations for soil moisture retrieval (Wang et al., 2018, RSE; Wang et al., 2019, JAG) and its use in closing water and energy budgets in TP, as well as inference of subsurface parameters from radiometric observations.

A reflection is made on modeling land-atmosphere radiative and heat-water transfer processes as a key component of Earth System Model (Zhao et al., 2018, ESSD; Yu et al., 2018, JGR).

A specific investigation has also been conducted, as part of the ESM, on the turbulent flux and energy budget over a high-altitude lake on the Tibetan Plateau (Wang et al., 2015, 2017, JGR; Wang et al., 2018, TAC).

As part of capacity building, three PhD (will) have graduated at the University of Twente (Q. Wang, S. Lv and B. Wang) in 2018-2019 and five new students have contributed to different aspects of the project (J. Du, Y. Yu, P. Zhang, M. Li, S. Mwangi).

In the past one year, based on in-situ measurements, reanalysis data, satellite remote sensing and numerical modeling, several main achievements have been acquired to promote the understanding of water and energy cycles over the Tibetan Plateau (TP). (1) Based on geostationary and polar orbiting satellite data, the surface energy balance system (SEBS) was used to derive hourly land surface heat fluxes at a spatial resolution of 10 km. Six stations scattered through the TP and equipped for flux tower measurements were used to perform cross-validation. The results showed good agreement between derived fluxes and in situ measurements through 3738 validation samples. The RMSEs for net radiation flux, sensible heat flux, latent heat flux and soil heat flux were 76.63 Wm^-2, 60.29 Wm^-2, 71.03 Wm^-2 and 37.5 Wm^-2, respectively. The derived results were also found to be superior to GLDAS flux products. (2) Based on field albedo measurements, moderate resolution imaging spectrometer (MODIS) albedo products and numerical simulation, we evaluated the ice albedo parameterization schemes in existing lake models and investigate the characteristics of the ice surface albedo in six typical TP lakes, as well as the influence of ice albedo error in the FLake model. Compared with observations, several ice albedo schemes all clearly overestimate the lake ice albedo by 0.26 to 0.66, while the average bias of MODIS albedo products is only 0.07. The MODIS-observed albedo of a snow-covered lake varies with the snow proportion, and the lake surface albedo in a snow-free state is approximately 0.15 during the frozen period. The simulated lake surface temperature is sensitive to variations in lake ice albedo especially in the spring and winter. (3) The climatological characteristics of water vapor and its interannual variability over the TP were investigated by using the ERA-interim monthly mean reanalysis datasets from 1979 to 2014. The analyses show that the TP is a water vapor convergence area, where the convergence was enhanced from 1979 to 2014. The TP is a moisture sink at a climatological mean, with an annual average net water vapor flux of . Detailed features in the water vapor flux and transport changes across the TP’s four boundaries were explored by simulating backward trajectories with a Lagrangian trajectory model (hybrid single-particle Lagrangian integrated trajectory model, HYSPLIT). In the study period, the water vapor contribution rate of midlatitude westerlies to the eastern and southern boundaries increased. However, the South Asian monsoon’s water vapor contribution decreased. (4) Six numerical experiments using the Weather Research and Forecasting (WRF) model were conducted to simulate a snow event over the TP in March 2017. The best performance was achieved when applying the CLM land surface physics. A potentially important factor is the advanced parameterization of albedo in CLM. The WRF model has a certain advantage to identify the snow event, but poor performance on accurate snow depth simulation; Sensitivity of total solid precipitation over the entire event to the land surface physics was larger than to the initial and boundary conditions. (5) Training of young scientists in the area of climate and environment. Six PhD students have been sent to European partner for joint training. Three of them have got PhD degree in University of Twente under the supervision of European PI (Prof. Zhongbo Su) and Chinese PI (Prof. Yaoming Ma). Several European students from our partner also come to China regularly for joint field visiting and academic exchange.

It is well known that the distribution of cloud and precipitation is affected by atmospheric parameters such as water vapor and updraft motion, as well as by topography. Due to the high altitude of Qinghai-Xizang Plateau (QXP) impacting on cloud and precipitation associated with latent heat release, lots of little known characteristics of them has been revealed continuously based on the measurements of Precipitation Radar (PR) onboard the Tropical Rainfall Measurement Mission (TRMM) satellite. In this study, the characteristics of the vertical structure of precipitation on QXP were studied and compared with the surrounding areas based on the 15 years' measurements of the PR. The results show that, firstly, there is no obvious brightness band in the vertical structure of precipitation over QXP, but it occurs in vertical structure measured by ground based precipitation radar that has the vertical resolution of 33 meters. The thickness of the brightness band is so thin that it is speculated that the PR with vertical resolution of 250 meters is not enough to distinguish the brightness band of precipitation over QXP. Secondly, according to the characteristics of sounding temperature and humidity profile, the precipitation in QXP can be divided into three types: deep strong convection, deep weak convection and shallow convection The Statistical calculations show that the precipitation over QXP is mainly in the form of week deep convection, which occupies 67. 8% followed by the form of shallow precipitation with 26. 4% and the strong deep convection with 5. 8%. Thirdly, the precipitation frequency peaks of deep strong convection and deep weak convection occurred at 16:00 (local time, the same below), and the precipitation intensity peaks for both types at 18:00 and 13:00, respectively. There is second peak at 0:00 for deep strong convection. The peak of precipitation frequency and intensity for shallow precipitation appeared at 20:00, which showed the characteristics of night rain. Finally, the shape of the average profile of deep strong convective precipitation over the QXT is similar to that of deep convective precipitation on the mainland of the Mid-East China, but different from that on the ocean surface.

Key Words: Precipitation, TRMM PR, Vertical Structure, Diurnal Variation

Surface soil moisture (SSM) plays a significant role in the water and energy fluxes at the land-atmosphere interface and its spatiotemporal dynamics is of crucial importance for e.g. water resources and agricultural management. In recent decades, with the development of remote sensing technologies, regional/global soil moisture estimation has been proceeding rapidly. The validation of soil moisture products from remote sensing is normally conducted by in-situ measurements directly or by interpolated soil moisture. However, the problem of mismatching scales between satellite imagery and ground-based observations typically imposes a large range of uncertainty in assessing the accuracy of remote sensing based soil moisture products. Interpolation schemes to upscale soil moisture from point measurements are a suitable alternative option of validating remote sensing soil moisture retrievals. This study aims to evaluate four upscaling methods (Inverse Distance Weighting (IDW), Ordinary Kriging (OK), Universal Kriging (UK) and Spline function (S)) individually on analysing a specific data source over a typical study area (Maqu in the north-eastern Qinghai-Tibet Plateau, HiWATER in the Heihe river basin and REMEDHUS in the semi-arid parts of the Duero Basin of Spain). The strengths and weaknesses of four upscaling methods are explained in detail by cross validation to enumerate all possibilities, thereby maximally reducing the generation of error. In summary, OK and UK show two best upscaling results at similar level, while IDW is inferior to them. S is the least suitable method for soil moisture upscaling, showing weak performance. As expected, a dense soil moisture network with a high number of stations is an important factor for a reliable upscaling of soil moisture. It is found that at least 30 stations (5.5 km × 5.5 km square area) over HiWATER are required to accurately compute geostatistical algorithms (UK, OK or IDW), among which OK performs best in upscaling soil moisture. In cases with less than 30 stations, it’s not recommended to apply geostatistical algorithms to upscale soil moisture, especially for Maqu and REMEDHUS, two research areas which are widely used for soil moisture validation.

With combined remote sensing observations in the microwave- and optical- region of the electromagnetic spectrum we intend to study the dynamics of an alpine meadow floral ecosystem. The microwave observations are made with a broadband (1 – 10 GHz) full polarimetric ground-based scatterometer. From these observations we want to retrieve the soil moisture content, above-ground biomass, and Leaf Area Index. Installed next to the scatterometer is a dual field-of-view (upwelling and downwelling) high resolution (up to 0.1 nm) spectrometer system covering the 400 – 900 nm range. From the spectrometer observations we want to retrieve chlorophyll fluorescence vegetation water content, and pigment factions. Both observations are done over multiple month periods with one measurement every hour.

During the conference we will present a poster on the experimental approach, the technical specifications, and measured observations of the scatterometer. The scatterometer setup was designed to be simple consisting of commercial off-the -shelf components: a Vector Network Analyser and two dual polarization broadband antennas, one for transmit, the other for receive. Also, the experimental approach for the long-term observations entails the two antennas fixed in one position, so that an (automated) rotational stage was not necessary.

Due to the broadband approach the antennas used have broad Gain patterns, especially for the lower frequencies. This property, together with the antenna’s close proximity to the ground (5 m above the surface) requires that for the derivation of the backscattering coefficient (s⁰) the ground surface mapping of G²/R⁴ must be evaluated to deduce the scatterometer’s footprint and associated angle of incidence interval.

To reduce the fading-induced uncertainty of s⁰ we employ frequency averaging techniques. In order to select proper bandwidths over which to average we use predictions of the frequency- and angle dependent behaviour of s⁰ calculated with the Advanced Integral Equation model (I²Em).

We will present an analysis on the spatial variability of s⁰ at our measurement site. Data for this analysis were obtained by measuring the s⁰ over different antenna elevation- and azimuth angles. Additionally, we show the temporal behaviour of the measured s⁰ during some of the key seasonal moments: frozen soil with senescent vegetation, thawing period, spring period (low vegetation) and summer period (maximum vegetation biomass).

L-band passive 1 microwave remote sensing is one of the most effective methods to map the global soil moisture distribution, yet, at which soil depth satellites are measuring is still inconclusive. Recently, with the Lv’s multilayer soil effective temperature scheme, such depth information can be revealed in the framework of the zeroth-order incoherent model when soil temperature varies linearly with soil optical depth. In this paper, we examine the relationships between soil temperature microwave sensing depth, penetration depth, and soil effective temperature, considering the nonlinear case. The soil temperature sensing depth often also named penetration depth is redefined as the depth where soil temperature equals the soil effective temperature. A method is developed to estimate soil temperature sensing depth from one pair of soil temperature and moisture measurement at an arbitrary depth, the soil surface temperature, and the deep soil temperature which is assumed to be constant in time. The method can be used to estimate the soil effective temperature and soil temperature sensing depth.

With combined remote sensing observations in the microwave- and optical- region of the electromagnetic spectrum we intend to study the dynamics of an alpine meadow floral ecosystem. The microwave observations are made with a broadband (1 – 10 GHz) full polarimetric ground-based scatterometer. From these observations we want to retrieve the soil moisture content, above-ground biomass, and Leaf Area Index. Installed next to the scatterometer is a dual field-of-view (upwelling and downwelling) high resolution (up to 0.1 nm) spectrometer system covering the 400 – 900 nm range. From the spectrometer observations we want to retrieve chlorophyll fluorescence vegetation water content, and pigment factions. Both observations are done over multiple month periods with one measurement every hour.

During the conference we will present a poster on the experimental approach, the technical specifications, and measured observations of the scatterometer. The scatterometer setup was designed to be simple consisting of commercial off-the -shelf components: a Vector Network Analyser and two dual polarization broadband antennas, one for transmit, the other for receive. Also, the experimental approach for the long-term observations entails the two antennas fixed in one position, so that an (automated) rotational stage was not necessary.

Due to the broadband approach the antennas used have broad Gain patterns, especially for the lower frequencies. This property, together with the antenna’s close proximity to the ground (5 m above the surface) requires that for the derivation of the backscattering coefficient (s⁰) the ground surface mapping of G²/R⁴ must be evaluated to deduce the scatterometer’s footprint and associated angle of incidence interval.

To reduce the fading-induced uncertainty of s⁰ we employ frequency averaging techniques. In order to select proper bandwidths over which to average we use predictions of the frequency- and angle dependent behaviour of s⁰ calculated with the Advanced Integral Equation model (I²Em).

We will present an analysis on the spatial variability of s⁰ at our measurement site. Data for this analysis were obtained by measuring the s⁰ over different antenna elevation- and azimuth angles. Additionally, we show the temporal behaviour of the measured s⁰ during some of the key seasonal moments: frozen soil with senescent vegetation, thawing period, spring period (low vegetation) and summer period (maximum vegetation biomass).