Actual land parcels vary significantly in size and complexity. Previous studies were limited by existing technical methods for fine-grained land use classification. The emergence of multimodal large language models offers new techniques for image classification, but their application in land use classification remains unexplored. This study presents LandGPT, a multimodal large language model trained on the CN-MSLU-100K dataset, covering fine-grained land use classification of irregular parcels. This study proposes a trans-level discrimination framework to improve LandGPT’s ability to classify fine-grained land use. Under this framework, LandGPT achieves a discrimination accuracy of 89.7% and a Kappa coefficient of 0.85 for fine-grained land use categories, outperforming state-of-the-art models with a 48.33% accuracy improvement. In some challenging categories, the improvement reaches nearly 1500%. This study finds that training with multi-source remote sensing image data improved LandGPT’s accuracy by 15.79% compared to single-image data. This study explores Prompt engineering based on LandGPT. The optimal prompt paradigm offers fine-grained categories and guides the model for accurate classification, reducing errors from LLM hallucinations. This study pioneeringly explores the application of large language models in the land use domain and offers a new solution for fine-grained land use classification.

The spatial representativeness error of in situ soil moisture (SM) is recognized as a major source of uncertainty when validating satellite SM products with a spatial resolution of tens of kilometers. Site underrepresentation is primarily caused by environmental heterogeneity, but their relationship remains poorly understood. Here, we assessed the spatial representativeness of in situ SM from 322 strictly screened stations worldwide relative to coarse-resolution (~0.25°) satellite footprint based on the extended triple collocation (ETC) method. We then evaluated the influence of the heterogeneity of four environmental factors (soil texture, land cover types, elevation, and vegetation coverage) on site representativeness. Moreover, we calculated SM variability within the satellite footprint based on 1-km SM data to explore its relationship with environmental heterogeneity. Results indicate that about 63% of the sites have relatively good spatial representativeness (ETC-derived correlation coefficient$\ge 0.7$). Soil texture and land cover exhibit greater heterogeneity across the mid and high latitudes of the Northern Hemisphere. The larger heterogeneity in elevation and vegetation coverage is primarily found in regions with significant ridges and dense vegetation, respectively. Land cover is the major factor influencing the spatial representativeness of SM sites, and the increase in the heterogeneity of land cover enhances SM variability, which negatively impacts site representativeness. The in situ SM can be more representative when the proportion of the land cover type where the site is located is higher or when there are fewer land cover types within the satellite footprint. Moreover, it is found that the newly proposed metric of the similar area ratio of sites, as a measure of land cover heterogeneity, can effectively reflect SM variability. This metric can also serve as a supplementary criterion for selecting representative sites, particularly in situations where sites are sparse and the ETC method is inapplicable. These findings provide useful references for robust evaluation of satellite SM products based on in situ measurements (e.g., in situ SM upscaling and SM site deployment).

The comprehensive and robust assessment of diverse global-scale satellite-based soil moisture products from various satellite data sources (e.g., different frequencies and incidence angles) and retrieval algorithms is essential for the refinements as well as applications of these products. To date, soil moisture retrieval algorithms and products are rapidly evolving and their updated iterations are ongoing. In support of the validation activities of recently developed/reprocessed satellite soil moisture products, the study firstly assessed eight commonly-employed satellite soil moisture datasets comprising SMAP (DCA, IB, and MTDCA), SMOS-IC, AMSR2 (LPRM and JAXA), FY-3C, and ESA CCI on a global scale using three different strategies, i.e., ERA5 reanalysis soil moisture dataset with similar spatial resolution to satellite products,in situmeasurements from densely-instrumented networks worldwide with mitigated spatial mismatch between ground site and satellite pixel, and the Extended Triple Collocation (ETC) method that can obtain error indicators relative to ground truth. The skills of these products under a broad range of vegetation density, land cover and climate types, and surface heterogeneity (heterogeneity in terrain, land cover, soil texture, and vegetation coverage) were also examined. The results indicate: (1) different soil moisture products show overall consistency in skill ranking under three different evaluation strategies, except for SMAP DCA, SMAP-IB, and SMAP MTDCA in terms ofRvalue; (2) ESA CCI, SMAP-IB, SMAP DCA products generally perform better than the others under three strategies, and SMOS-IC and SMAP MTDCA also show satisfactory performance concerning ubRMSD andRvalues; (3) vegetation density exerts visible influences on satellite soil moisture datasets. Specifically, the C/X-band (AMSR2 and FY-3C) and L-band (SMAP and SMOS) products display the optimal skills under sparse and moderate vegetation coverage respectively, and the impacts of vegetation density on C/X-band products are evidently stronger than those on L-band datasets. The errors of satellite soil moisture data also increase as the increase of heterogeneity in terrain, land cover, and vegetation coverage, while the effect of heterogeneity in soil texture on the skill of satellite soil moisture products is insignificant; (4) the skills of L-band products are more stable than those of C/X-band datasets under different ground conditions.

Climatology (mean seasonal cycle) often dominates the errors in satellite soil moisture (SM) products, which is highly essential for the water-carbon cycle in tropical rainforests. Although microwave observations at L-band are expected to provide more accurate SM information benefiting from their stronger penetration capacity, the SM mapping in rainforests by L-band measurements is still challenging and is less investigated by the community. To bridge the research gap, five L-band satellite SM products from the Soil Moisture and Ocean Salinity (SMOS) and Soil Moisture Active Passive (SMAP) satellites, including SMOS-IC, SMAP SCA-V, DCA, MTDCA and IB were assessed by using the FLUXNET SM data in tropical rainforests. To cope with the time inconsistence between satellite and ground data, the SM climatology variations in rainforests for diverse time periods were compared using long-term ERA5 SM. The results indicate the SM climatology is relatively stable over different time periods during 2001-2020 for rainforest sites. Based on the stability of SM climatology, L-band SM products were demonstrated to satisfactorily capture the SM climatology variations in rainforests, especially for SMOS-IC and SMAP-IB. The results are expected to provide guidelines for the hydro-ecological applications using satellite SM products in tropical rainforests.

Blending both active and passive microwave measurements are expected to provide more robust surface soil moisture (SSM) estimations over various environmental conditions compared to that from the single sensor. The integration of the newest L-band passive (i.e., Soil Moisture Active Passive, SMAP) with the similar observation scale active (i.e., the Advanced Scatterometer, ASCAT) sensors provides a considerable opportunity to improve the accuracy of SSM mapping, which however is rarely investigated to date. In this study, we implemented the integration of SMAP brightness temperature (TB) and ASCAT backscattering coefficient (σ) for estimating SSM using the machine learning approach, by fully considering the error sources in physically-based retrieval approaches (e.g., τ–ω model). The independent validation results using ground data from 14 dense networks show the integration of SMAP and ASCAT measurements can satisfactorily achieve better SSM retrievals compared to ASCAT SSM, SMAP-SCA-V SSM and ESA CCI SSM products, with the lowest ubRMSE of 0.042 m3/m3and the highest R of 0.76. This study is expected to enrich the understanding of SSM retrieval from active and passive microwave satellites, and provide SSM product with higher accuracy for eco-hydrological applications.

Assessment of the model-based surface soil temperature (ST) products is very important for hydrometeorological and ecological applications, as well as model refinements. Distinguished from previous regional validations using only in situ observations from sparse networks, this study focused on the evaluation of model-based ST products by considering ground observations from 15 dense networks worldwide from April 2015 to December 2017 covering a wide range of ground conditions. Four model-based ST products were selected for the assessment, including the Modern-Era Retrospective Analysis for Research and Applications, version 2 (MERRA-2), the Goddard Earth Observing System Model version 5 Forward Processing (GEOS-5 FP), the ERA-Interim and its successor, the newly developed ERA5. The results indicate the GEOS-5 ST product slightly outperforms other ST products by showing an averaged ubRMSD of 1.84 K. All model-based ST products underestimate in situ ST with a negative bias. All four model-based ST products are demonstrated to well capture the temporal trends of ground observations with very promising$R$values larger than 0.97. The ERA5 shows visible improvements compared to its predecessor ERA-Interim by exhibiting smaller ubRMSD, absolute bias and larger$R$values. These findings are expected to provide useful suggestions for the enhancement and specific usage of the model-based ST products.

Soil moisture is an essential variable both in environmental monitoring research and application. With the requirement of high-precision soil moisture data, it is highly necessary to construct in-situ soil moisture sensors Web in high density, in which the economics, complexity, and low-power consumption of sensor Webs should be essentially considered. However, most current existing soil moisture monitoring networks have limitations due to high power consumption, complex architecture, and expensive equipment. These issues are not conducive to high-density soil moisture observation. In view of the existing problems in the current soil moisture in-situ sites, an effective resolution for high-density soil moisture observation is needed. For the first time, we are bringing Narrow-Band Internet of Things (NB-IoT), a low-power Internet-of-Things technology, into geospatial sensor Web for soil moisture observation. Moreover, we built a high-density in-situ soil moisture sensor Web via NB-IoT and compared it with ZigBee at the Baoxie experimental zone in Wuhan, China, there about 1 km2, and we acquired data for up to 12 months. We analyze the acquisition record of battery status, signals, and soil moisture. We analyzed the battery life, the impact of the signal on the battery performance, the signal quality, and data acquisition clearly and intuitively. This unprecedented study and application, we discovered and concluded that the low-power sensor Web of the NB-IoT communication protocol could be suitable for high-density soil moisture observation.

Land use and cover change (LUCC) is a central component in current strategies for managing natural resources and monitoring environmental change. In this paper, we used maximum likelihood classification algorithm to obtain the supervised land use and cover classification. Four major land use and cover classes are identified and mapped from 2000 to 2015. The changes of land use and cover using Landsat images of the study area were analyzed. The results showed that: From 2005 to 2015, closed forest has increased and the annual rate of change was (+)3.3%. Open forest has an annual rate of change of (+)1.21%. Water bodies had an annual rate of change of (+)0.81%. While the settlements and bare lands had a decrease of 52.93 km2and the annual rate of change was (-)5.3%.

Drought is one of the main natural hazards affecting the environment and economy of countries all over the world. Fusing weather data with satellite images therefore becomes a superior method of identifying and monitoring drought in a given region. We established the relationship between land surface temperature (LST), the normalized differential vegetation index (NDVI) and rainfall data to derive areas of drought. Then, we obtained the indexes from the rainfall anomaly and NDVI anomaly as indicators which confirm the drought indicative claims of the maps produced. Our further examination of the NDVI, LST and rainfall maps indicate that the western, central and Volta Regions of the study area are the least prone to drought, with Axim (one of the most southern towns) in Ghana recording the highest rainfall in the country each year.

Early warning of cyanobacteria bloom is very important for water ecological management of lakes. Thus, Based on Landsat8 OLI data, combining the trophic state index (TSI), cyanobacteria and macrophytes index (CMI) and floating algae index (FAI), we proposed a risk zoning framework of cyanobacteria bloom, and applied it in Lake Longgan (China). Eutrophication frequency of Lake Longgan has worsened since 2017. But the maximum eutrophication proportion occurred on 16 February 2016. With a deeper analysis on 16 February 2016, we find that the risk area of cyanobacteria bloom in Lake Longgan is distributed horizontally from the east bank to the center of the lake, and mostly is mid risk. High risk areas are mainly distributed in the east coast, showing a southwest direction. Therefore, the proposed risk zoning framework provides a useful approach to obtain the pre-warning information of cyanobacteria bloom in some Eutrophic and macrophytic lakes.

Drought is a weather phenomenon widespread worldwide due to the water shortage or unbalance of supply and demand, and it's also one of the most serious natural disasters for human life and agricultural production. The middle reach of Yangtze river, one of China's most important grain producer, subjected to the sub-tropical monsoon climate, is prone to have droughts. This paper has practical implications as it build a model by depending on the normalized difference vegetation index (NDVI) and land surface temperature (LST) of moderate resolution imaging spectroradiometer (MODIS) between 2005 and 2009. Firstly, the 8-day LST and 16-day NDVI data, 8-day LST and 30-day NDVI data were utilized to construct the LST/NDVI feature space. Secondly, the temperature vegetation dryness index (TVDI) images of the reach were derived respectively. Thirdly, the temporal evolution and spatial variation of drought was analyzed. Finally, the results of two different years were compared to analyze the drought in the reach. Our study showed the drought in May was more severe than that in other months. Therefore, a severe drought event is more likely to happen in May in the middle reach of Yangtze river and more measures should be taken to alleviate the loss for the governments.

Clouds often limit the ability of optical satellite sensors (such as the newly launched Gaofen-1 (GF-1) satellite in China) to observe regional soil moisture at high spatial resolutions, especially under full-cloud-contamination condition. Thus, accurate reconstruction of regional soil moisture in this case became a great methodological challenge because of the complexity and ill-posed nature of the problem. In this paper, we present a Satellite and In situ sensor Collaborated Reconstruction (SICR) method. In this method, four reconstruction rules were proposed to rebuild four kinds of corresponding missing pixels, defined as follows: C1 pixel (including one in situ sensor in its area), C2 pixel (physically similar to C1), C3 pixel (with a regular soil moisture observation sequence), and C4 pixel (remaining). By analyzing soil moisture observation relationships between these four types of pixels with in situ measurements and within these pixels, four numerical reconstruction rules were established. Linear regression, similar pixel determination, least square method, and geostatistical interpolation algorithms were used in these four rules. At last, all blank soil moisture pixels in the target soil moisture image can be filled by the SICR method. The experiment conducted in the central south of U.S. integrated 11 in situ soil moisture sensors from the United States Department of Agriculture with 11 GF-1 satellite soil moisture images. It was demonstrated that GF-1 soil moisture observations on October 17, 2014, were successfully reconstructed by the SICR method, based on the evaluations of visual appearance comparison, error distribution analysis, subimage comparison, average relative error, and universal image quality index. SICR also performed better than the reconstruction results only based on in situ or satellite sensor data. Moreover, the comparison with the soil moisture observation from the microwave sensor demonstrated the value of SICR in regional high-resolution soil moisture reconstruction. It was suggested that the SICR method provided an effective reconstruction method under full cloud contamination and showed great potential for collaborating satellite and in situ sensors.

Under the indoor building environment, when the fires and other accidents occur, how to effectively organize the masses evacuation and fire rescue, is closely related to the safety of people's lives and property and has become a critical problem of public concern. This paper presents an improved ant colony algorithm (ACO) to solve the problem of how to optimize the evacuation route and rescue route when an accident occurs. According to the key factors affecting people emergency evacuation, such as indoor building environment, fire and its combustion products, problem of path's optimal selection, etc., we propose an emergency evacuation model, based on the model it can give an optimal evacuation route for the mass and an optimal rescue route for the firefighters. We also analyzes the search results, it shows that the search results is robust and reasonable.

In China, drought is one of the major environmental disasters, which bring great harm to the people. The middle reach of Yangtze River is the most important base to produce grains in China. Influenced by the summer monsoon, the drought occurs frequently. In our paper, the NDVI and LST from MODIS data were utilized to calculate the TVDI (Temperature Vegetation Dryness Index), which were used to monitor the drought of the study area. Meteorological drought indices were calculated from 10-day precipitation, temperature and evaporation data of 94 meteorological stations, including precipitation standardized variables, dryness and relative moisture index were used to analyze the degree of drought and the area of drought. The results showed that TVDI is significantly related to soil moisture.

Different Earth-observing (EO) sensors have various capabilities for diverse observing tasks. Sensor planning services make the choice of web-ready sensors for specific observing tasks with regard to observing requests and sensor capabilities. Sensor capabilities rely on various parameters; thus, choosing EO sensors for specific observing tasks relying directly on these parameters is a multicriteria decision process. A sensor's capability can be drawn from these parameters with the help of an algorithm. Furthermore, if divided into different clusters based on capabilities, applicable sensors can be more easily chosen for a category of observing tasks. In this paper, a spaceborne EO optical sensor static capability index (SSCI) mechanism is drawn from an evaluation-and-clustering algorithm, which is composed of a self-organizing neural map in combination with weighted principal component analysis. The scheme of SSCI relies on no expert analysis system and thus is more flexible and efficient. EO scenarios of disaster reactions are among the application of this algorithm. In particular, scenarios of flooding disaster forecasting, relief aiding, and postdisaster loss assessment within the framework of International Charter on Space and Major Disasters have been utilized for experiments. They have shown that the SSCI assessing algorithm is feasible and stable, and the EO optical sensor clustering algorithm based on SSCI can offer reasonable clustering accuracies of EO optical sensors. In our experiments, the EO optical sensor SSCI computation and clustering algorithm had a time consumption within 2 s and 2 min, respectively, and memory consumption within 200 MB on a normal personal computer.

The sensor web is a new method in the earth observation field, comprising sensors that can sense, compute, and correspond with the World Wide Web. The Chinese Integrated Earth Observation Sensor Web (CIEOSW) project is a five-year national basic research program conducted by Wuhan University since 2011 and is aimed to address the following major challenges: (1) the lack of collaboration within the satellite observation system, (2) the absence of a coupling mechanism in heterogeneous spaceborne -- airborne -- ground sensors, and (3) the limited connection between monitoring and decision-support services. For two years, the CIEOSW project has strived to achieve the following: (1) a theory on the coupling and modeling of the Earth observation sensor web (EOSW), (2) an event-driven multi-sensor collaborative observation method, (3) an EOSW fusion and assimilation method, (4) an EOSW-based information extraction and rapid change detection method, (5) a task-oriented EOSW focusing service model, and (6) an epicontinental environment observation and analysis of typical areas in China.